diff options
| author | Jules Laplace <julescarbon@gmail.com> | 2019-04-17 22:08:59 +0200 |
|---|---|---|
| committer | Jules Laplace <julescarbon@gmail.com> | 2019-04-17 22:08:59 +0200 |
| commit | fba670e97b1baee6739aacf55325ce8dfd835be5 (patch) | |
| tree | c5a88f9964c8cc87a22331128580750c5f874a7b | |
| parent | 699d7a77b9d4120dfb75f271cb924b0e05a2fcaa (diff) | |
| parent | 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 (diff) | |
merge
18 files changed, 129 insertions, 122 deletions
diff --git a/site/content/pages/datasets/brainwash/assets/00425000_640x480.jpg b/site/content/pages/datasets/brainwash/assets/00425000_640x480.jpg Binary files differdeleted file mode 100644 index de62175a..00000000 --- a/site/content/pages/datasets/brainwash/assets/00425000_640x480.jpg +++ /dev/null diff --git a/site/content/pages/datasets/brainwash/assets/00425000_960.jpg b/site/content/pages/datasets/brainwash/assets/00425000_960.jpg Binary files differdeleted file mode 100644 index caa96fe2..00000000 --- a/site/content/pages/datasets/brainwash/assets/00425000_960.jpg +++ /dev/null diff --git a/site/content/pages/datasets/brainwash/assets/brainwash_example.jpg b/site/content/pages/datasets/brainwash/assets/brainwash_example.jpg Binary files differnew file mode 100755 index 00000000..3ddf4323 --- /dev/null +++ b/site/content/pages/datasets/brainwash/assets/brainwash_example.jpg diff --git a/site/content/pages/datasets/brainwash/assets/brainwash_grid.jpg b/site/content/pages/datasets/brainwash/assets/brainwash_grid.jpg Binary files differnew file mode 100755 index 00000000..0271ec6a --- /dev/null +++ b/site/content/pages/datasets/brainwash/assets/brainwash_grid.jpg diff --git a/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay.jpg b/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay.jpg Binary files differdeleted file mode 100755 index 2f5917e3..00000000 --- a/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay.jpg +++ /dev/null diff --git a/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay_wm.jpg b/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay_wm.jpg Binary files differdeleted file mode 100755 index 790dbb79..00000000 --- a/site/content/pages/datasets/brainwash/assets/brainwash_mean_overlay_wm.jpg +++ /dev/null diff --git a/site/content/pages/datasets/brainwash/assets/brainwash_montage.jpg b/site/content/pages/datasets/brainwash/assets/brainwash_montage.jpg Binary files differdeleted file mode 100644 index 193fdd03..00000000 --- a/site/content/pages/datasets/brainwash/assets/brainwash_montage.jpg +++ /dev/null diff --git a/site/content/pages/datasets/brainwash/index.md b/site/content/pages/datasets/brainwash/index.md index 156b02c7..b57bcdf4 100644 --- a/site/content/pages/datasets/brainwash/index.md +++ b/site/content/pages/datasets/brainwash/index.md @@ -19,30 +19,24 @@ authors: Adam Harvey ### sidebar ### end sidebar -*Brainwash* is a head detection dataset created from San Francisco's Brainwash Cafe livecam footage. It includes 11,918 images of "everyday life of a busy downtown cafe"[^readme] captured at 100 second intervals throught the entire day. Brainwash dataset was captured during 3 days in 2014: October 27, November 13, and November 24. According the author's reserach paper introducing the dataset, the images were acquired with the help of Angelcam.com.[^end_to_end] +Brainwash is a dataset of livecam images taken from San Francisco's Brainwash Cafe. It includes 11,918 images of "everyday life of a busy downtown cafe"[^readme] captured at 100 second intervals throught the entire day. The Brainwash dataset includes 3 full days of webcam images taken on October 27, November 13, and November 24 in 2014. According the author's [reserach paper](https://www.semanticscholar.org/paper/End-to-End-People-Detection-in-Crowded-Scenes-Stewart-Andriluka/1bd1645a629f1b612960ab9bba276afd4cf7c666) introducing the dataset, the images were acquired with the help of Angelcam.com[^end_to_end] -Brainwash is not a widely used dataset but since its publication by Stanford University in 2015, it has notably appeared in several research papers from the National University of Defense Technology in Changsha, China. In 2016 and in 2017 researchers there conducted studies on detecting people's heads in crowded scenes for the purpose of surveillance. [^localized_region_context] [^replacement_algorithm] +The Brainwash dataset is unique because it uses images from a publicly available webcam that records people inside a privately owned business without any consent. No ordinary cafe custom could ever suspect there image would end up in dataset used for surveillance reserach and development, but that is exactly what happened to customers at Brainwash cafe in San Francisco. -If you happen to have been at Brainwash cafe in San Francisco at any time on October 26, November 13, or November 24 in 2014 you are most likely included in the Brainwash dataset and have unwittingly contributed to surveillance research. +Although Brainwash appears to be a less popular dataset, it was used in 2016 and 2017 by researchers from the National University of Defense Technology in China took note of the dataset and used it for two [research](https://www.semanticscholar.org/paper/Localized-region-context-and-object-feature-fusion-Li-Dou/b02d31c640b0a31fb18c4f170d841d8e21ffb66c) [projects](https://www.semanticscholar.org/paper/A-Replacement-Algorithm-of-Non-Maximum-Suppression-Zhao-Wang/591a4bfa6380c9fcd5f3ae690e3ac5c09b7bf37b) on advancing the capabilities of object detection to more accurately isolate the target region in an image ([PDF](https://www.itm-conferences.org/articles/itmconf/pdf/2017/04/itmconf_ita2017_05006.pdf)). [^localized_region_context] [^replacement_algorithm]. The dataset also appears in a 2017 [research paper](https://ieeexplore.ieee.org/document/7877809) from Peking University for the purpose of improving surveillance capabilities for "people detection in the crowded scenes". -{% include 'dashboard.html' %} -{% include 'supplementary_header.html' %} + - +{% include 'dashboard.html' %} - +{% include 'supplementary_header.html' %} - + -TODO + -- change supp images to 2x2 grid with bboxes -- add bounding boxes to the header image -- remake montage with randomized images, with bboxes -- add ethics link to Stanford -- add optout info {% include 'cite_our_work.html' %} @@ -52,4 +46,4 @@ TODO [^readme]: "readme.txt" https://exhibits.stanford.edu/data/catalog/sx925dc9385. [^end_to_end]: Stewart, Russel. Andriluka, Mykhaylo. "End-to-end people detection in crowded scenes". 2016. [^localized_region_context]: Li, Y. and Dou, Y. and Liu, X. and Li, T. Localized Region Context and Object Feature Fusion for People Head Detection. ICIP16 Proceedings. 2016. Pages 594-598. -[^replacement_algorithm]: Zhao. X, Wang Y, Dou, Y. A Replacement Algorithm of Non-Maximum Suppression Base on Graph Clustering.
\ No newline at end of file +[^replacement_algorithm]: Zhao. X, Wang Y, Dou, Y. A Replacement Algorithm of Non-Maximum Suppression Base on Graph Clustering. diff --git a/site/content/pages/datasets/duke_mtmc/index.md b/site/content/pages/datasets/duke_mtmc/index.md index 2140fed7..0f4986de 100644 --- a/site/content/pages/datasets/duke_mtmc/index.md +++ b/site/content/pages/datasets/duke_mtmc/index.md @@ -22,7 +22,7 @@ Duke MTMC (Multi-Target, Multi-Camera) is a dataset of surveillance video footag In this investigation into the Duke MTMC dataset we tracked down over 100 publicly available research papers that explicitly acknowledged using Duke MTMC. Our analysis shows that the dataset has spread far beyond its origins and intentions in academic research projects at Duke University. Since its publication in 2016, more than twice as many research citations originated in China as in the United States. Among these citations were papers with explicit and direct links to the Chinese military and several of the companies known to provide Chinese authorities with the oppressive surveillance technology used to monitor millions of Uighur Muslims. -In one 2018 [paper](http://openaccess.thecvf.com/content_cvpr_2018/papers/Xu_Attention-Aware_Compositional_Network_CVPR_2018_paper.pdf) jointly published by researchers from SenseNets and SenseTime (and funded by SenseTime Group Limited) entitled [Attention-Aware Compositional Network for Person Re-identification](https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e), the Duke MTMC dataset was used for "extensive experiments" on improving person re-identification across multiple surveillance cameras with important applications in "finding missing elderly and children, and suspect tracking, etc." Both SenseNets and SenseTime have been directly linked to the providing surveillance technology to monitor Uighur Muslims in China. [^sensetime_qz][^sensenets_uyghurs][^xinjiang_nyt] +In one 2018 [paper](http://openaccess.thecvf.com/content_cvpr_2018/papers/Xu_Attention-Aware_Compositional_Network_CVPR_2018_paper.pdf) jointly published by researchers from SenseNets and SenseTime (and funded by SenseTime Group Limited) entitled [Attention-Aware Compositional Network for Person Re-identification](https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e), the Duke MTMC dataset was used for "extensive experiments" on improving person re-identification across multiple surveillance cameras with important applications in "finding missing elderly and children, and suspect tracking, etc." Both SenseNets and SenseTime have been directly linked to the providing surveillance technology to monitor Uighur Muslims in China. [^xinjiang_nyt][^sensetime_qz][^sensenets_uyghurs]  @@ -30,18 +30,17 @@ Despite [repeated](https://www.hrw.org/news/2017/11/19/china-police-big-data-sys | Organization | Paper | Link | Year | Used Duke MTMC | |---|---|---|---| -| SenseNets, SenseTime | Attention-Aware Compositional Network for Person Re-identification | [SemanticScholar](https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e) | 2018 | ✔ | -|SenseTime| End-to-End Deep Kronecker-Product Matching for Person Re-identification | [thcvf.com](http://openaccess.thecvf.com/content_cvpr_2018/papers/Shen_End-to-End_Deep_Kronecker-Product_CVPR_2018_paper.pdf) | 2018| ✔ | -|CloudWalk| Horizontal Pyramid Matching for Person Re-identification | [arxiv.org](https://arxiv.org/pdf/1804.05275.pdf) | 20xx | ✔ | -| Megvii | Multi-Target, Multi-Camera Tracking by Hierarchical Clustering: Recent Progress on DukeMTMC Project | [SemanticScholar](https://www.semanticscholar.org/paper/Multi-Target%2C-Multi-Camera-Tracking-by-Hierarchical-Zhang-Wu/10c20cf47d61063032dce4af73a4b8e350bf1128) | 2018 | ✔ | +| Beihang University | Orientation-Guided Similarity Learning for Person Re-identification | [ieee.org](https://ieeexplore.ieee.org/document/8545620) | 2018 | ✔ | +| Beihang University | Online Inter-Camera Trajectory Association Exploiting Person Re-Identification and Camera Topology | [acm.org](https://dl.acm.org/citation.cfm?id=3240663) | 2018 | ✔ | +| CloudWalk | CloudWalk re-identification technology extends facial biometric tracking with improved accuracy | [BiometricUpdate.com](https://www.biometricupdate.com/201903/cloudwalk-re-identification-technology-extends-facial-biometric-tracking-with-improved-accuracy) | 2019 | ✔ | +|CloudWalk| Horizontal Pyramid Matching for Person Re-identification | [arxiv.org](https://arxiv.org/pdf/1804.05275.pdf) | 2018 | ✔ | | Megvii | Person Re-Identification (slides) | [github.io](https://zsc.github.io/megvii-pku-dl-course/slides/Lecture%2011,%20Human%20Understanding_%20ReID%20and%20Pose%20and%20Attributes%20and%20Activity%20.pdf) | 2017 | ✔ | +| Megvii | Multi-Target, Multi-Camera Tracking by Hierarchical Clustering: Recent Progress on DukeMTMC Project | [SemanticScholar](https://www.semanticscholar.org/paper/Multi-Target%2C-Multi-Camera-Tracking-by-Hierarchical-Zhang-Wu/10c20cf47d61063032dce4af73a4b8e350bf1128) | 2018 | ✔ | | Megvii | SCPNet: Spatial-Channel Parallelism Network for Joint Holistic and Partial PersonRe-Identification | [arxiv.org](https://arxiv.org/abs/1810.06996) | 2018 | ✔ | -| CloudWalk | CloudWalk re-identification technology extends facial biometric tracking with improved accuracy | [BiometricUpdate.com](https://www.biometricupdate.com/201903/cloudwalk-re-identification-technology-extends-facial-biometric-tracking-with-improved-accuracy) | 2018 | ✔ | -| CloudWalk | Horizontal Pyramid Matching for Person Re-identification | [arxiv.org](https://arxiv.org/abs/1804.05275)] | 2018 | ✔ | | National University of Defense Technology | Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers | [SemanticScholar.org](https://www.semanticscholar.org/paper/Tracking-by-Animation%3A-Unsupervised-Learning-of-He-Liu/e90816e1a0e14ea1e7039e0b2782260999aef786) | 2018 | ✔ | | National University of Defense Technology | Unsupervised Multi-Object Detection for Video Surveillance Using Memory-Based Recurrent Attention Networks | [SemanticScholar.org](https://www.semanticscholar.org/paper/Unsupervised-Multi-Object-Detection-for-Video-Using-He-He/59f357015054bab43fb8cbfd3f3dbf17b1d1f881) | 2018 | ✔ | -| Beihang University | Orientation-Guided Similarity Learning for Person Re-identification | [ieee.org](https://ieeexplore.ieee.org/document/8545620) | 2018 | ✔ | -| Beihang University | Online Inter-Camera Trajectory Association Exploiting Person Re-Identification and Camera Topology | [acm.org](https://dl.acm.org/citation.cfm?id=3240663) | 2018 | ✔ | +| SenseNets, SenseTime | Attention-Aware Compositional Network for Person Re-identification | [SemanticScholar](https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e) | 2018 | ✔ | +|SenseTime| End-to-End Deep Kronecker-Product Matching for Person Re-identification | [thcvf.com](http://openaccess.thecvf.com/content_cvpr_2018/papers/Shen_End-to-End_Deep_Kronecker-Product_CVPR_2018_paper.pdf) | 2018| ✔ | The reasons that companies in China use the Duke MTMC dataset for research are technically no different than the reasons it is used in the United States and Europe. In fact, the original creators of the dataset published a follow up report in 2017 titled [Tracking Social Groups Within and Across Cameras](https://www.semanticscholar.org/paper/Tracking-Social-Groups-Within-and-Across-Cameras-Solera-Calderara/9e644b1e33dd9367be167eb9d832174004840400) with specific applications to "automated analysis of crowds and social gatherings for surveillance and security applications". Their work, as well as the creation of the original dataset in 2014 were both supported in part by the United States Army Research Laboratory. @@ -49,12 +48,12 @@ Citations from the United States and Europe show a similar trend to that in Chin | Organization | Paper | Link | Year | Used Duke MTMC | |---|---|---|---| -| IARPA, IBM, CloudWalk | Horizontal Pyramid Matching for Person Re-identification | [arxiv.org](https://arxiv.org/abs/1804.05275) | 2018 | ✔ | +| IARPA, IBM | Horizontal Pyramid Matching for Person Re-identification | [arxiv.org](https://arxiv.org/abs/1804.05275) | 2018 | ✔ | | Microsoft | ReXCam: Resource-Efficient, Cross-CameraVideo Analytics at Enterprise Scale | [arxiv.org](https://arxiv.org/abs/1811.01268) | 2018 | ✔ | | Microsoft | Scaling Video Analytics Systems to Large Camera Deployments | [arxiv.org](https://arxiv.org/pdf/1809.02318.pdf) | 2018 | ✔ | -| University College of London, National University of Defense Technology | Unsupervised Multi-Object Detection for Video Surveillance Using Memory-Based RecurrentAttention Networks | [SemanticScholar.org](https://pdfs.semanticscholar.org/59f3/57015054bab43fb8cbfd3f3dbf17b1d1f881.pdf) | 2018 | ✔ | -| Vision Semantics Ltd. | Unsupervised Person Re-identification by Deep Learning Tracklet Association | [arxiv.org](https://arxiv.org/abs/1809.02874) | 2018 | ✔ | +| University College of London | Unsupervised Multi-Object Detection for Video Surveillance Using Memory-Based RecurrentAttention Networks | [SemanticScholar.org](https://pdfs.semanticscholar.org/59f3/57015054bab43fb8cbfd3f3dbf17b1d1f881.pdf) | 2018 | ✔ | | US Dept. of Homeland Security | Re-Identification with Consistent Attentive Siamese Networks | [arxiv.org](https://arxiv.org/abs/1811.07487/) | 2019 | ✔ | +| Vision Semantics Ltd. | Unsupervised Person Re-identification by Deep Learning Tracklet Association | [arxiv.org](https://arxiv.org/abs/1809.02874) | 2018 | ✔ | By some metrics the dataset is considered a huge success. It is regarded as highly influential research and has contributed to hundreds, if not thousands, of projects to advance artificial intelligence for person tracking and monitoring. All the above citations, regardless of which country is using it, align perfectly with the original [intent](http://vision.cs.duke.edu/DukeMTMC/) of the Duke MTMC dataset: "to accelerate advances in multi-target multi-camera tracking". @@ -66,7 +65,7 @@ The same logic applies for all the new extensions of the Duke MTMC dataset inclu But this perspective comes at significant cost to civil rights, human rights, and privacy. The creation and distribution of the Duke MTMC illustrates an egregious prioritization of surveillance technologies over individual rights, where the simple act of going to class could implicate your biometric data in a surveillance training dataset, perhaps even used by foreign defense agencies against your own ethics, against your own political interests, or against universal human rights. -For the approximately 2,000 students in Duke MTMC dataset, there is unfortunately no escape. It would be impossible to remove oneself from all copies of the dataset downloaded around the world. Instead, over 2,000 students and visitors who happened to be walking to class on March 13, 2014 will forever remain in all downloaded copies of the Duke MTMC dataset and all its extensions, contributing to a global supply chain of data that powers governmental and commercial expansion of biometric surveillance technologies. +For the approximately 2,000 students in Duke MTMC dataset there is unfortunately no escape. It would be impossible to remove oneself from all copies of the dataset downloaded around the world. Instead, over 2,000 students and visitors who happened to be walking to class in 2014 will forever remain in all downloaded copies of the Duke MTMC dataset and all its extensions, contributing to a global supply chain of data that powers governmental and commercial expansion of biometric surveillance technologies.  @@ -80,7 +79,7 @@ For the approximately 2,000 students in Duke MTMC dataset, there is unfortunatel #### Video Timestamps -The video timestamps contain the likely, but not yet confirmed, date and times of capture. Because the video timestamps align with the start and stop [time sync data](http://vision.cs.duke.edu/DukeMTMC/details.html#time-sync) provided by the researchers, it at least aligns the relative time. The [rainy weather](https://www.wunderground.com/history/daily/KIGX/date/2014-3-19?req_city=Durham&req_state=NC&req_statename=North%20Carolina&reqdb.zip=27708&reqdb.magic=1&reqdb.wmo=99999) on that day also contributes towards the likelihood of March 14, 2014. +The video timestamps contain the likely, but not yet confirmed, date and times the video recorded. Because the video timestamps align with the start and stop [time sync data](http://vision.cs.duke.edu/DukeMTMC/details.html#time-sync) provided by the researchers, it at least confirms the relative timing. The [precipitous weather](https://www.wunderground.com/history/daily/KIGX/date/2014-3-19?req_city=Durham&req_state=NC&req_statename=North%20Carolina&reqdb.zip=27708&reqdb.magic=1&reqdb.wmo=99999) on March 14, 2014 in Durham, North Carolina supports, but does not confirm, that this day is a potential capture date. === columns 2 @@ -105,7 +104,13 @@ The video timestamps contain the likely, but not yet confirmed, date and times o #### Errata -- The Duke MTMC dataset paper mentions 2,700 identities, but their ground truth file only lists annotations for 1,812. +The original Duke MTMC dataset paper mentions 2,700 identities, but their ground truth file only lists annotations for 1,812, and their own research typically mentions 2,000. For this writeup we used 2,000 to describe the approximate number of students. + +#### Ethics + +Please direct any questions about the ethics of the dataset to Duke University's [Institutional Ethics & Compliance Office](https://hr.duke.edu/policies/expectations/compliance/) using the number at the bottom of the page. + +{% include 'cite_our_work.html' %} #### Citing Duke MTMC @@ -120,15 +125,10 @@ If you use any data from the Duke MTMC, please follow their [license](http://vis } </pre> -{% include 'cite_our_work.html' %} - -#### ToDo - -- clean up citations, formatting - ### Footnotes [^duke_mtmc_orig]: "Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking". 2016. [SemanticScholar](https://www.semanticscholar.org/paper/Performance-Measures-and-a-Data-Set-for-Tracking-Ristani-Solera/27a2fad58dd8727e280f97036e0d2bc55ef5424c) [^sensetime_qz]: <https://qz.com/1248493/sensetime-the-billion-dollar-alibaba-backed-ai-company-thats-quietly-watching-everyone-in-china/> [^sensenets_uyghurs]: <https://foreignpolicy.com/2019/03/19/962492-orwell-china-socialcredit-surveillance/> [^xinjiang_nyt]: Mozur, Paul. "One Month, 500,000 Face Scans: How China Is Using A.I. to Profile a Minority". https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html. April 14, 2019. + diff --git a/site/content/pages/datasets/index.md b/site/content/pages/datasets/index.md index c0373d60..289aa2fd 100644 --- a/site/content/pages/datasets/index.md +++ b/site/content/pages/datasets/index.md @@ -13,4 +13,4 @@ sync: false # Facial Recognition Datasets -Explore publicly available facial recognition datasets. More datasets will be added throughout 2019. +Explore publicly available facial recognition datasets feeding into research and development of biometric surveillance technologies at the largest technology companies and defense contractors in the world. diff --git a/site/content/pages/datasets/msceleb/index.md b/site/content/pages/datasets/msceleb/index.md index 70e85699..d5e52952 100644 --- a/site/content/pages/datasets/msceleb/index.md +++ b/site/content/pages/datasets/msceleb/index.md @@ -19,7 +19,6 @@ authors: Adam Harvey ### sidebar ### end sidebar -[ PAGE UNDER DEVELOPMENT ] https://www.hrw.org/news/2019/01/15/letter-microsoft-face-surveillance-technology diff --git a/site/content/pages/datasets/uccs/assets/uccs_grid.jpg b/site/content/pages/datasets/uccs/assets/uccs_grid.jpg Binary files differindex d3d898ea..95dff617 100644 --- a/site/content/pages/datasets/uccs/assets/uccs_grid.jpg +++ b/site/content/pages/datasets/uccs/assets/uccs_grid.jpg diff --git a/site/content/pages/datasets/uccs/index.md b/site/content/pages/datasets/uccs/index.md index 68fff4db..b6073384 100644 --- a/site/content/pages/datasets/uccs/index.md +++ b/site/content/pages/datasets/uccs/index.md @@ -20,43 +20,37 @@ authors: Adam Harvey ### sidebar ### end sidebar -UnConstrained College Students (UCCS) is a dataset of long-range surveillance photos captured at University of Colorado Colorado Springs developed primarily for research and development of "face detection and recognition research towards surveillance applications"[^uccs_vast]. According to the authors of two papers associated with the dataset, over 1,700 students and pedestrians were "photographed using a long-range high-resolution surveillance camera without their knowledge".[^funding_uccs] In this investigation, we examine the contents of the dataset, funding sources, photo EXIF data, and information from publicly available research project citations. - +UnConstrained College Students (UCCS) is a dataset of long-range surveillance photos captured at University of Colorado Colorado Springs developed primarily for research and development of "face detection and recognition research towards surveillance applications"[^uccs_vast]. According to the authors of [two](https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745) [papers](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1) associated with the dataset, over 1,700 students and pedestrians were "photographed using a long-range high-resolution surveillance camera without their knowledge".[^funding_uccs] In this investigation, we examine the contents of the [dataset](http://vast.uccs.edu/Opensetface/), its funding sources, photo EXIF data, and information from publicly available research project citations. The UCCS dataset includes over 1,700 unique identities, most of which are students walking to and from class. As of 2018, it was the "largest surveillance [face recognition] benchmark in the public domain."[^surv_face_qmul] The photos were taken during the spring semesters of 2012 – 2013 on the West Lawn of the University of Colorado Colorado Springs campus. The photographs were timed to capture students during breaks between their scheduled classes in the morning and afternoon during Monday through Thursday. "For example, a student taking Monday-Wednesday classes at 12:30 PM will show up in the camera on almost every Monday and Wednesday."[^sapkota_boult]. - + -The long-range surveillance images in the UnContsrained College Students dataset were captured using a Canon 7D 18 megapixel digital camera fitted with a Sigma 800mm F5.6 EX APO DG HSM telephoto lens and pointed out an office window across the university's West Lawn. The students were photographed from a distance of approximately 150 meters through an office window. "The camera [was] programmed to start capturing images at specific time intervals between classes to maximize the number of faces being captured."[^sapkota_boult] -Their setup made it impossible for students to know they were being photographed, providing the researchers with realistic surveillance images to help build face detection and recognition systems for real world applications in defense, intelligence, and commercial applications. - +The long-range surveillance images in the UnContsrained College Students dataset were taken using a Canon 7D 18-megapixel digital camera fitted with a Sigma 800mm F5.6 EX APO DG HSM telephoto lens and pointed out an office window across the university's West Lawn. The students were photographed from a distance of approximately 150 meters through an office window. "The camera [was] programmed to start capturing images at specific time intervals between classes to maximize the number of faces being captured."[^sapkota_boult] +Their setup made it impossible for students to know they were being photographed, providing the researchers with realistic surveillance images to help build face recognition systems for real world applications for defense, intelligence, and commercial partners. -In the two papers associated with the release of the UCCS dataset ([Unconstrained Face Detection and Open-Set Face Recognition Challenge](https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745) and [Large Scale Unconstrained Open Set Face Database](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1)), the researchers disclosed their funding sources as ODNI (United States Office of Director of National Intelligence), IARPA (Intelligence Advance Research Projects Activity), ONR MURI (Office of Naval Research and The Department of Defense Multidisciplinary University Research Initiative), Army SBIR (Small Business Innovation Research), SOCOM SBIR (Special Operations Command and Small Business Innovation Research), and the National Science Foundation. Further, UCCS's VAST site explicity [states](https://vast.uccs.edu/project/iarpa-janus/) they are part of the [IARPA Janus](https://www.iarpa.gov/index.php/research-programs/janus), a face recognition project developed to serve the needs of national intelligence interests. + -The EXIF data embedded in the images shows that the photo capture times follow a similar pattern, but also highlights that the vast majority of photos (over 7,000) were taken on Tuesdays around noon during students' lunch break. The lack of any photos taken on Friday shows that the researchers were only interested in capturing images of students. +The EXIF data embedded in the images shows that the photo capture times follow a similar pattern to that outlined by the researchers, but also highlights that the vast majority of photos (over 7,000) were taken on Tuesdays around noon during students' lunch break. The lack of any photos taken between Friday through Sunday shows that the researchers were only interested in capturing images of students during the peak campus hours.   -The two research papers associated with the release of the UCCS dataset ([Unconstrained Face Detection and Open-Set Face Recognition Challenge](https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745) and [Large Scale Unconstrained Open Set Face Database](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1)), acknowledge that the primary funding sources for their work were United States defense and intelligence agencies. Specifically, development of the UnContrianed College Students dataset was funded by the Intelligence Advanced Research Projects Activity (IARPA), Office of Director of National Intelligence (ODNI), Office of Naval Research and The Department of Defense Multidisciplinary University Research Initiative (ONR MURI), Small Business Innovation Research (SBIR), Special Operations Command and Small Business Innovation Research (SOCOM SBIR), and the National Science Foundation. Further, UCCS's VAST site explicitly [states](https://vast.uccs.edu/project/iarpa-janus/) they are part of the [IARPA Janus](https://www.iarpa.gov/index.php/research-programs/janus), a face recognition project developed to serve the needs of national intelligence interests, clearly establishing the the funding sources and immediate benefactors of this dataset are United States defense and intelligence agencies. - - -Although the images were first captured in 2012 – 2013 the dataset was not publicly released until 2016. Then in 2017 the UCCS face dataset formed the basis for a defense and intelligence agency funded [face recognition challenge](http://www.face-recognition-challenge.com/) project at the International Joint Biometrics Conference in Denver, CO. And in 2018 the dataset was again used for the [2nd Unconstrained Face Detection and Open Set Recognition Challenge](https://erodner.github.io/ial2018eccv/) at the European Computer Vision Conference (ECCV) in Munich, Germany. - -As of April 15, 2019, the UCCS dataset is no longer available for public download. But during the three years it was publicly available (2016-2019) the UCCS dataset appeared in at least 6 publicly available research papers including verified usage from Beihang University who is known to provide research and development for China's military. +The two research papers associated with the release of the UCCS dataset ([Unconstrained Face Detection and Open-Set Face Recognition Challenge](https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745) and [Large Scale Unconstrained Open Set Face Database](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1)), acknowledge that the primary funding sources for their work were United States defense and intelligence agencies. Specifically, development of the UnContsrianed College Students dataset was funded by the Intelligence Advanced Research Projects Activity (IARPA), Office of Director of National Intelligence (ODNI), Office of Naval Research and The Department of Defense Multidisciplinary University Research Initiative (ONR MURI), and the Special Operations Command and Small Business Innovation Research (SOCOM SBIR) amongst others. UCCS's VAST site also explicitly [states](https://vast.uccs.edu/project/iarpa-janus/) their involvement in the [IARPA Janus](https://www.iarpa.gov/index.php/research-programs/janus) face recognition project developed to serve the needs of national intelligence, establishing that immediate benefactors of this dataset include United States defense and intelligence agencies, but it would go on to benefit other similar organizations. +In 2017, one year after its public release, the UCCS face dataset formed the basis for a defense and intelligence agency funded [face recognition challenge](http://www.face-recognition-challenge.com/) project at the International Joint Biometrics Conference in Denver, CO. And in 2018 the dataset was again used for the [2nd Unconstrained Face Detection and Open Set Recognition Challenge](https://erodner.github.io/ial2018eccv/) at the European Computer Vision Conference (ECCV) in Munich, Germany. +As of April 15, 2019, the UCCS dataset is no longer available for public download. But during the three years it was publicly available (2016-2019) the UCCS dataset appeared in at least 6 publicly available research papers including verified usage from Beihang University who is known to provide research and development for China's military; and Vision Semantics Ltd who lists the UK Ministory of Defence as a project partner. {% include 'dashboard.html' %} {% include 'supplementary_header.html' %} - -To show the types of face images used in the UCCS student dataset while protecting their individual privacy, a generative adversarial network was used to interpolate between identities in the dataset. The image below shows a generative adversarial network trained on the UCCS face bounding box areas from 16,000 images and over 90,000 face regions. +Since this site To show the types of face images used in the UCCS student dataset while protecting their individual privacy, a generative adversarial network was used to interpolate between identities in the dataset. The image below shows a generative adversarial network trained on the UCCS face bounding box areas from 16,000 images and over 90,000 face regions.  @@ -98,7 +92,7 @@ To show the types of face images used in the UCCS student dataset while protecti ### Location -The location of the camera and subjects can confirmed using several visual cues in the dataset images: the unique pattern of the sidewalk that is only used on the UCCS Pedestrian Spine near the West Lawn, the two UCCS sign poles with matching graphics still visible in Google Street View, the no parking sign and directionality of its arrow, the back of street sign next to it, the slight bend in the sidewalk, the presence of cars passing in the background of the image, and the far wall of the parking garage all match images in the dataset. The [original papers](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1) also provides another clue: a [picture of the camera](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1/figure/1) inside the office that was used to create the dataset. The window view in this image provides another match for the brick pattern on the north facade of the Kraember Family Library and the green metal fence along the sidewalk. View the [location on Google Maps](https://www.google.com/maps/place/University+of+Colorado+Colorado+Springs/@38.8934297,-104.7992445,27a,35y,258.51h,75.06t/data=!3m1!1e3!4m5!3m4!1s0x87134fa088fe399d:0x92cadf3962c058c4!8m2!3d38.8968312!4d-104.8049528) +The location of the camera and subjects was confirmed using several visual cues in the dataset images: the unique pattern of the sidewalk that is only used on the UCCS Pedestrian Spine near the West Lawn, the two UCCS sign poles with matching graphics still visible in Google Street View, the no parking sign and directionality of its arrow, the back of street sign next to it, the slight bend in the sidewalk, the presence of cars passing in the background of the image, and the far wall of the parking garage all match images in the dataset. The [original papers](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1) also provides another clue: a [picture of the camera](https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1/figure/1) inside the office that was used to create the dataset. The window view in this image provides another match for the brick pattern on the north facade of the Kraember Family Library and the green metal fence along the sidewalk. View the [location on Google Maps](https://www.google.com/maps/place/University+of+Colorado+Colorado+Springs/@38.8934297,-104.7992445,27a,35y,258.51h,75.06t/data=!3m1!1e3!4m5!3m4!1s0x87134fa088fe399d:0x92cadf3962c058c4!8m2!3d38.8968312!4d-104.8049528)  diff --git a/site/public/datasets/brainwash/index.html b/site/public/datasets/brainwash/index.html index becc8949..0f782924 100644 --- a/site/public/datasets/brainwash/index.html +++ b/site/public/datasets/brainwash/index.html @@ -49,11 +49,10 @@ </div><div class='meta'> <div class='gray'>Website</div> <div><a href='https://purl.stanford.edu/sx925dc9385' target='_blank' rel='nofollow noopener'>stanford.edu</a></div> - </div></div><p><em>Brainwash</em> is a head detection dataset created from San Francisco's Brainwash Cafe livecam footage. It includes 11,918 images of "everyday life of a busy downtown cafe"<a class="footnote_shim" name="[^readme]_1"> </a><a href="#[^readme]" class="footnote" title="Footnote 1">1</a> captured at 100 second intervals throught the entire day. Brainwash dataset was captured during 3 days in 2014: October 27, November 13, and November 24. According the author's reserach paper introducing the dataset, the images were acquired with the help of Angelcam.com.<a class="footnote_shim" name="[^end_to_end]_1"> </a><a href="#[^end_to_end]" class="footnote" title="Footnote 2">2</a></p> -<p>People's Liberation Army National University of Defense Science and Technology</p> -<p>Brainwash is not a widely used dataset but since its publication by Stanford University in 2015, it has notably appeared in several research papers from the National University of Defense Technology in Changsha, China. In 2016 and in 2017 researchers there conducted studies on detecting people's heads in crowded scenes for the purpose of surveillance. <a class="footnote_shim" name="[^localized_region_context]_1"> </a><a href="#[^localized_region_context]" class="footnote" title="Footnote 3">3</a> <a class="footnote_shim" name="[^replacement_algorithm]_1"> </a><a href="#[^replacement_algorithm]" class="footnote" title="Footnote 4">4</a></p> -<p>If you happen to have been at Brainwash cafe in San Francisco at any time on October 26, November 13, or November 24 in 2014 you are most likely included in the Brainwash dataset and have unwittingly contributed to surveillance research.</p> -</section><section> + </div></div><p>Brainwash is a dataset of livecam images taken from San Francisco's Brainwash Cafe. It includes 11,918 images of "everyday life of a busy downtown cafe"<a class="footnote_shim" name="[^readme]_1"> </a><a href="#[^readme]" class="footnote" title="Footnote 1">1</a> captured at 100 second intervals throught the entire day. The Brainwash dataset includes 3 full days of webcam images taken on October 27, November 13, and November 24 in 2014. According the author's <a href="https://www.semanticscholar.org/paper/End-to-End-People-Detection-in-Crowded-Scenes-Stewart-Andriluka/1bd1645a629f1b612960ab9bba276afd4cf7c666">reserach paper</a> introducing the dataset, the images were acquired with the help of Angelcam.com<a class="footnote_shim" name="[^end_to_end]_1"> </a><a href="#[^end_to_end]" class="footnote" title="Footnote 2">2</a></p> +<p>The Brainwash dataset is unique because it uses images from a publicly available webcam that records people inside a privately owned business without any consent. No ordinary cafe custom could ever suspect there image would end up in dataset used for surveillance reserach and development, but that is exactly what happened to customers at Brainwash cafe in San Francisco.</p> +<p>Although Brainwash appears to be a less popular dataset, it was used in 2016 and 2017 by researchers from the National University of Defense Technology in China took note of the dataset and used it for two <a href="https://www.semanticscholar.org/paper/Localized-region-context-and-object-feature-fusion-Li-Dou/b02d31c640b0a31fb18c4f170d841d8e21ffb66c">research</a> <a href="https://www.semanticscholar.org/paper/A-Replacement-Algorithm-of-Non-Maximum-Suppression-Zhao-Wang/591a4bfa6380c9fcd5f3ae690e3ac5c09b7bf37b">projects</a> on advancing the capabilities of object detection to more accurately isolate the target region in an image (<a href="https://www.itm-conferences.org/articles/itmconf/pdf/2017/04/itmconf_ita2017_05006.pdf">PDF</a>). <a class="footnote_shim" name="[^localized_region_context]_1"> </a><a href="#[^localized_region_context]" class="footnote" title="Footnote 3">3</a> <a class="footnote_shim" name="[^replacement_algorithm]_1"> </a><a href="#[^replacement_algorithm]" class="footnote" title="Footnote 4">4</a>. The dataset also appears in a 2017 <a href="https://ieeexplore.ieee.org/document/7877809">research paper</a> from Peking University for the purpose of improving surveillance capabilities for "people detection in the crowded scenes".</p> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/brainwash_grid.jpg' alt=' A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc'><div class='caption'> A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc</div></div></section><section> <h3>Who used Brainwash Dataset?</h3> <p> @@ -113,15 +112,7 @@ <h2>Supplementary Information</h2> -</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/brainwash_saliency_map.jpg' alt=' A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc'><div class='caption'> A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/00425000_960.jpg' alt=' An sample image from the Brainwash dataset used for training face and head detection algorithms for surveillance. The datset contains about 12,000 images. License: Open Data Commons Public Domain Dedication (PDDL)'><div class='caption'> An sample image from the Brainwash dataset used for training face and head detection algorithms for surveillance. The datset contains about 12,000 images. License: Open Data Commons Public Domain Dedication (PDDL)</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/brainwash_montage.jpg' alt=' 49 of the 11,918 images included in the Brainwash dataset. License: Open Data Commons Public Domain Dedication (PDDL)'><div class='caption'> 49 of the 11,918 images included in the Brainwash dataset. License: Open Data Commons Public Domain Dedication (PDDL)</div></div></section><section><p>TODO</p> -<ul> -<li>change supp images to 2x2 grid with bboxes</li> -<li>add bounding boxes to the header image</li> -<li>remake montage with randomized images, with bboxes</li> -<li>add ethics link to Stanford</li> -<li>add optout info</li> -</ul> -</section><section> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/brainwash_example.jpg' alt=' An sample image from the Brainwash dataset used for training face and head detection algorithms for surveillance. The datset contains about 12,000 images. License: Open Data Commons Public Domain Dedication (PDDL)'><div class='caption'> An sample image from the Brainwash dataset used for training face and head detection algorithms for surveillance. The datset contains about 12,000 images. License: Open Data Commons Public Domain Dedication (PDDL)</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/brainwash/assets/brainwash_saliency_map.jpg' alt=' A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc'><div class='caption'> A visualization of 81,973 head annotations from the Brainwash dataset training partition. © megapixels.cc</div></div></section><section> <h4>Cite Our Work</h4> <p> @@ -134,7 +125,7 @@ title = {MegaPixels: Origins, Ethics, and Privacy Implications of Publicly Available Face Recognition Image Datasets}, year = 2019, url = {https://megapixels.cc/}, - urldate = {2019-04-20} + urldate = {2019-04-18} }</pre> </p> diff --git a/site/public/datasets/duke_mtmc/index.html b/site/public/datasets/duke_mtmc/index.html index 86cccd11..3c0bc0c2 100644 --- a/site/public/datasets/duke_mtmc/index.html +++ b/site/public/datasets/duke_mtmc/index.html @@ -48,7 +48,11 @@ <div><a href='http://vision.cs.duke.edu/DukeMTMC/' target='_blank' rel='nofollow noopener'>duke.edu</a></div> </div></div><p>Duke MTMC (Multi-Target, Multi-Camera) is a dataset of surveillance video footage taken on Duke University's campus in 2014 and is used for research and development of video tracking systems, person re-identification, and low-resolution facial recognition. The dataset contains over 14 hours of synchronized surveillance video from 8 cameras at 1080p and 60FPS with over 2 million frames of 2,000 students walking to and from classes. The 8 surveillance cameras deployed on campus were specifically setup to capture students "during periods between lectures, when pedestrian traffic is heavy"<a class="footnote_shim" name="[^duke_mtmc_orig]_1"> </a><a href="#[^duke_mtmc_orig]" class="footnote" title="Footnote 1">1</a>.</p> <p>In this investigation into the Duke MTMC dataset we tracked down over 100 publicly available research papers that explicitly acknowledged using Duke MTMC. Our analysis shows that the dataset has spread far beyond its origins and intentions in academic research projects at Duke University. Since its publication in 2016, more than twice as many research citations originated in China as in the United States. Among these citations were papers with explicit and direct links to the Chinese military and several of the companies known to provide Chinese authorities with the oppressive surveillance technology used to monitor millions of Uighur Muslims.</p> +<<<<<<< HEAD <p>In one 2018 <a href="http://openaccess.thecvf.com/content_cvpr_2018/papers/Xu_Attention-Aware_Compositional_Network_CVPR_2018_paper.pdf">paper</a> jointly published by researchers from SenseNets and SenseTime (and funded by SenseTime Group Limited) entitled <a href="https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e">Attention-Aware Compositional Network for Person Re-identification</a>, the Duke MTMC dataset was used for "extensive experiments" on improving person re-identification across multiple surveillance cameras with important applications in "finding missing elderly and children, and suspect tracking, etc." Both SenseNets and SenseTime have been directly linked to the providing surveillance technology to monitor Uighur Muslims in China. <a class="footnote_shim" name="[^sensetime_qz]_1"> </a><a href="#[^sensetime_qz]" class="footnote" title="Footnote 2">2</a><a class="footnote_shim" name="[^sensenets_uyghurs]_1"> </a><a href="#[^sensenets_uyghurs]" class="footnote" title="Footnote 3">3</a><a class="footnote_shim" name="[^xinjiang_nyt]_1"> </a><a href="#[^xinjiang_nyt]" class="footnote" title="Footnote 4">4</a></p> +======= +<p>In one 2018 <a href="http://openaccess.thecvf.com/content_cvpr_2018/papers/Xu_Attention-Aware_Compositional_Network_CVPR_2018_paper.pdf">paper</a> jointly published by researchers from SenseNets and SenseTime (and funded by SenseTime Group Limited) entitled <a href="https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e">Attention-Aware Compositional Network for Person Re-identification</a>, the Duke MTMC dataset was used for "extensive experiments" on improving person re-identification across multiple surveillance cameras with important applications in "finding missing elderly and children, and suspect tracking, etc." Both SenseNets and SenseTime have been directly linked to the providing surveillance technology to monitor Uighur Muslims in China. <a class="footnote_shim" name="[^xinjiang_nyt]_1"> </a><a href="#[^xinjiang_nyt]" class="footnote" title="Footnote 1">1</a><a class="footnote_shim" name="[^sensetime_qz]_1"> </a><a href="#[^sensetime_qz]" class="footnote" title="Footnote 2">2</a><a class="footnote_shim" name="[^sensenets_uyghurs]_1"> </a><a href="#[^sensenets_uyghurs]" class="footnote" title="Footnote 3">3</a></p> +>>>>>>> 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 </section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/duke_mtmc/assets/duke_mtmc_reid_montage.jpg' alt=' A collection of 1,600 out of the approximately 2,000 students and pedestrians in the Duke MTMC dataset. These students were also included in the Duke MTMC Re-ID dataset extension used for person re-identification, and eventually the QMUL SurvFace face recognition dataset. Open Data Commons Attribution License.'><div class='caption'> A collection of 1,600 out of the approximately 2,000 students and pedestrians in the Duke MTMC dataset. These students were also included in the Duke MTMC Re-ID dataset extension used for person re-identification, and eventually the QMUL SurvFace face recognition dataset. Open Data Commons Attribution License.</div></div></section><section><p>Despite <a href="https://www.hrw.org/news/2017/11/19/china-police-big-data-systems-violate-privacy-target-dissent">repeated</a> <a href="https://www.hrw.org/news/2018/02/26/china-big-data-fuels-crackdown-minority-region">warnings</a> by Human Rights Watch that the authoritarian surveillance used in China represents a violation of human rights, researchers at Duke University continued to provide open access to their dataset for anyone to use for any project. As the surveillance crisis in China grew, so did the number of citations with links to organizations complicit in the crisis. In 2018 alone there were over 70 research projects happening in China that publicly acknowledged benefiting from the Duke MTMC dataset. Amongst these were projects from SenseNets, SenseTime, CloudWalk, Megvii, Beihang University, and the PLA's National University of Defense Technology.</p> <table> <thead><tr> @@ -61,30 +65,30 @@ </thead> <tbody> <tr> -<td>SenseNets, SenseTime</td> -<td>Attention-Aware Compositional Network for Person Re-identification</td> -<td><a href="https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e">SemanticScholar</a></td> +<td>Beihang University</td> +<td>Orientation-Guided Similarity Learning for Person Re-identification</td> +<td><a href="https://ieeexplore.ieee.org/document/8545620">ieee.org</a></td> <td>2018</td> <td>✔</td> </tr> <tr> -<td>SenseTime</td> -<td>End-to-End Deep Kronecker-Product Matching for Person Re-identification</td> -<td><a href="http://openaccess.thecvf.com/content_cvpr_2018/papers/Shen_End-to-End_Deep_Kronecker-Product_CVPR_2018_paper.pdf">thcvf.com</a></td> +<td>Beihang University</td> +<td>Online Inter-Camera Trajectory Association Exploiting Person Re-Identification and Camera Topology</td> +<td><a href="https://dl.acm.org/citation.cfm?id=3240663">acm.org</a></td> <td>2018</td> <td>✔</td> </tr> <tr> <td>CloudWalk</td> -<td>Horizontal Pyramid Matching for Person Re-identification</td> -<td><a href="https://arxiv.org/pdf/1804.05275.pdf">arxiv.org</a></td> -<td>20xx</td> +<td>CloudWalk re-identification technology extends facial biometric tracking with improved accuracy</td> +<td><a href="https://www.biometricupdate.com/201903/cloudwalk-re-identification-technology-extends-facial-biometric-tracking-with-improved-accuracy">BiometricUpdate.com</a></td> +<td>2019</td> <td>✔</td> </tr> <tr> -<td>Megvii</td> -<td>Multi-Target, Multi-Camera Tracking by Hierarchical Clustering: Recent Progress on DukeMTMC Project</td> -<td><a href="https://www.semanticscholar.org/paper/Multi-Target%2C-Multi-Camera-Tracking-by-Hierarchical-Zhang-Wu/10c20cf47d61063032dce4af73a4b8e350bf1128">SemanticScholar</a></td> +<td>CloudWalk</td> +<td>Horizontal Pyramid Matching for Person Re-identification</td> +<td><a href="https://arxiv.org/pdf/1804.05275.pdf">arxiv.org</a></td> <td>2018</td> <td>✔</td> </tr> @@ -97,22 +101,15 @@ </tr> <tr> <td>Megvii</td> -<td>SCPNet: Spatial-Channel Parallelism Network for Joint Holistic and Partial PersonRe-Identification</td> -<td><a href="https://arxiv.org/abs/1810.06996">arxiv.org</a></td> -<td>2018</td> -<td>✔</td> -</tr> -<tr> -<td>CloudWalk</td> -<td>CloudWalk re-identification technology extends facial biometric tracking with improved accuracy</td> -<td><a href="https://www.biometricupdate.com/201903/cloudwalk-re-identification-technology-extends-facial-biometric-tracking-with-improved-accuracy">BiometricUpdate.com</a></td> +<td>Multi-Target, Multi-Camera Tracking by Hierarchical Clustering: Recent Progress on DukeMTMC Project</td> +<td><a href="https://www.semanticscholar.org/paper/Multi-Target%2C-Multi-Camera-Tracking-by-Hierarchical-Zhang-Wu/10c20cf47d61063032dce4af73a4b8e350bf1128">SemanticScholar</a></td> <td>2018</td> <td>✔</td> </tr> <tr> -<td>CloudWalk</td> -<td>Horizontal Pyramid Matching for Person Re-identification</td> -<td><a href="https://arxiv.org/abs/1804.05275">arxiv.org</a>]</td> +<td>Megvii</td> +<td>SCPNet: Spatial-Channel Parallelism Network for Joint Holistic and Partial PersonRe-Identification</td> +<td><a href="https://arxiv.org/abs/1810.06996">arxiv.org</a></td> <td>2018</td> <td>✔</td> </tr> @@ -131,16 +128,16 @@ <td>✔</td> </tr> <tr> -<td>Beihang University</td> -<td>Orientation-Guided Similarity Learning for Person Re-identification</td> -<td><a href="https://ieeexplore.ieee.org/document/8545620">ieee.org</a></td> +<td>SenseNets, SenseTime</td> +<td>Attention-Aware Compositional Network for Person Re-identification</td> +<td><a href="https://www.semanticscholar.org/paper/Attention-Aware-Compositional-Network-for-Person-Xu-Zhao/14ce502bc19b225466126b256511f9c05cadcb6e">SemanticScholar</a></td> <td>2018</td> <td>✔</td> </tr> <tr> -<td>Beihang University</td> -<td>Online Inter-Camera Trajectory Association Exploiting Person Re-Identification and Camera Topology</td> -<td><a href="https://dl.acm.org/citation.cfm?id=3240663">acm.org</a></td> +<td>SenseTime</td> +<td>End-to-End Deep Kronecker-Product Matching for Person Re-identification</td> +<td><a href="http://openaccess.thecvf.com/content_cvpr_2018/papers/Shen_End-to-End_Deep_Kronecker-Product_CVPR_2018_paper.pdf">thcvf.com</a></td> <td>2018</td> <td>✔</td> </tr> @@ -159,7 +156,7 @@ </thead> <tbody> <tr> -<td>IARPA, IBM, CloudWalk</td> +<td>IARPA, IBM</td> <td>Horizontal Pyramid Matching for Person Re-identification</td> <td><a href="https://arxiv.org/abs/1804.05275">arxiv.org</a></td> <td>2018</td> @@ -180,32 +177,38 @@ <td>✔</td> </tr> <tr> -<td>University College of London, National University of Defense Technology</td> +<td>University College of London</td> <td>Unsupervised Multi-Object Detection for Video Surveillance Using Memory-Based RecurrentAttention Networks</td> <td><a href="https://pdfs.semanticscholar.org/59f3/57015054bab43fb8cbfd3f3dbf17b1d1f881.pdf">SemanticScholar.org</a></td> <td>2018</td> <td>✔</td> </tr> <tr> -<td>Vision Semantics Ltd.</td> -<td>Unsupervised Person Re-identification by Deep Learning Tracklet Association</td> -<td><a href="https://arxiv.org/abs/1809.02874">arxiv.org</a></td> -<td>2018</td> -<td>✔</td> -</tr> -<tr> <td>US Dept. of Homeland Security</td> <td>Re-Identification with Consistent Attentive Siamese Networks</td> <td><a href="https://arxiv.org/abs/1811.07487/">arxiv.org</a></td> <td>2019</td> <td>✔</td> </tr> +<tr> +<td>Vision Semantics Ltd.</td> +<td>Unsupervised Person Re-identification by Deep Learning Tracklet Association</td> +<td><a href="https://arxiv.org/abs/1809.02874">arxiv.org</a></td> +<td>2018</td> +<td>✔</td> +</tr> </tbody> </table> <p>By some metrics the dataset is considered a huge success. It is regarded as highly influential research and has contributed to hundreds, if not thousands, of projects to advance artificial intelligence for person tracking and monitoring. All the above citations, regardless of which country is using it, align perfectly with the original <a href="http://vision.cs.duke.edu/DukeMTMC/">intent</a> of the Duke MTMC dataset: "to accelerate advances in multi-target multi-camera tracking".</p> +<<<<<<< HEAD <p>The same logic applies for all the new extensions of the Duke MTMC dataset including <a href="https://github.com/layumi/DukeMTMC-reID_evaluation">Duke MTMC Re-ID</a>, <a href="https://github.com/Yu-Wu/DukeMTMC-VideoReID">Duke MTMC Video Re-ID</a>, Duke MTMC Groups, and <a href="https://github.com/vana77/DukeMTMC-attribute">Duke MTMC Attribute</a>. And it also applies to all the new specialized datasets that will be created from Duke MTMC, such as the low-resolution face recognition dataset called <a href="https://qmul-survface.github.io/">QMUL-SurvFace</a>, which was funded in part by <a href="https://seequestor.com">SeeQuestor</a>, a computer vision provider to law enforcement agencies including Scotland Yards and Queensland Police. From the perspective of academic researchers, security contractors, and defense agencies using these datasets to advance their organization's work, Duke MTMC provides significant value regardless of who else is using it, so long as it advances their own interests in artificial intelligence.</p> </section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/duke_mtmc/assets/duke_mtmc_saliencies.jpg' alt=' Duke MTMC pedestrian detection saliency maps for 8 cameras deployed on campus © megapixels.cc'><div class='caption'> Duke MTMC pedestrian detection saliency maps for 8 cameras deployed on campus © megapixels.cc</div></div></section><section><p>But this perspective comes at significant cost to civil rights, human rights, and privacy. The creation and distribution of the Duke MTMC illustrates an egregious prioritization of surveillance technologies over individual rights, where the simple act of going to class could implicate your biometric data in a surveillance training dataset, perhaps even used by foreign defense agencies against your own ethics, against your own political interests, or against universal human rights.</p> <p>For the approximately 2,000 students in Duke MTMC dataset, there is unfortunately no escape. It would be impossible to remove oneself from all copies of the dataset downloaded around the world. Instead, over 2,000 students and visitors who happened to be walking to class on March 13, 2014 will forever remain in all downloaded copies of the Duke MTMC dataset and all its extensions, contributing to a global supply chain of data that powers governmental and commercial expansion of biometric surveillance technologies.</p> +======= +<p>The same logic applies for all the new extensions of the Duke MTMC dataset including <a href="https://github.com/layumi/DukeMTMC-reID_evaluation">Duke MTMC Re-ID</a>, <a href="https://github.com/Yu-Wu/DukeMTMC-VideoReID">Duke MTMC Video Re-ID</a>, Duke MTMC Groups, and <a href="https://github.com/vana77/DukeMTMC-attribute">Duke MTMC Attribute</a>. And it also applies to all the new specialized datasets that will be created from Duke MTMC, such as the low-resolution face recognition dataset called <a href="https://qmul-survface.github.io/">QMUL-SurvFace</a>, which was funded in part by <a href="https://seequestor.com">SeeQuestor</a>, a computer vision provider to law enforcement agencies including Scotland Yards and Queensland Police. From the perspective of academic researchers, security contractors, and defense agencies using these datasets to advance their organization's work, Duke MTMC provides significant value regardless of who else is using it so long as it accelerate advances their own interests in artificial intelligence.</p> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/duke_mtmc/assets/duke_mtmc_saliencies.jpg' alt=' Duke MTMC pedestrian detection saliency maps for 8 cameras deployed on campus © megapixels.cc'><div class='caption'> Duke MTMC pedestrian detection saliency maps for 8 cameras deployed on campus © megapixels.cc</div></div></section><section><p>But this perspective comes at significant cost to civil rights, human rights, and privacy. The creation and distribution of the Duke MTMC illustrates an egregious prioritization of surveillance technologies over individual rights, where the simple act of going to class could implicate your biometric data in a surveillance training dataset, perhaps even used by foreign defense agencies against your own ethics, against universal human rights, or against your own political interests.</p> +<p>For the approximately 2,000 students in Duke MTMC dataset there is unfortunately no escape. It would be impossible to remove oneself from all copies of the dataset downloaded around the world. Instead, over 2,000 students and visitors who happened to be walking to class in 2014 will forever remain in all downloaded copies of the Duke MTMC dataset and all its extensions, contributing to a global supply chain of data that powers governmental and commercial expansion of biometric surveillance technologies.</p> +>>>>>>> 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 </section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/duke_mtmc/assets/duke_mtmc_cameras.jpg' alt=' Duke MTMC camera views for 8 cameras deployed on campus © megapixels.cc'><div class='caption'> Duke MTMC camera views for 8 cameras deployed on campus © megapixels.cc</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/duke_mtmc/assets/duke_mtmc_camera_map.jpg' alt=' Duke MTMC camera locations on Duke University campus. Open Data Commons Attribution License.'><div class='caption'> Duke MTMC camera locations on Duke University campus. Open Data Commons Attribution License.</div></div></section><section> <h3>Who used Duke MTMC Dataset?</h3> @@ -267,7 +270,11 @@ <h2>Supplementary Information</h2> </section><section><h4>Video Timestamps</h4> +<<<<<<< HEAD <p>The video timestamps contain the likely, but not yet confirmed, date and times of capture. Because the video timestamps align with the start and stop <a href="http://vision.cs.duke.edu/DukeMTMC/details.html#time-sync">time sync data</a> provided by the researchers, it at least aligns the relative time. The <a href="https://www.wunderground.com/history/daily/KIGX/date/2014-3-19?req_city=Durham&req_state=NC&req_statename=North%20Carolina&reqdb.zip=27708&reqdb.magic=1&reqdb.wmo=99999">rainy weather</a> on that day also contributes towards the likelihood of March 14, 2014.</p> +======= +<p>The video timestamps contain the likely, but not yet confirmed, date and times the video recorded. Because the video timestamps align with the start and stop <a href="http://vision.cs.duke.edu/DukeMTMC/details.html#time-sync">time sync data</a> provided by the researchers, it at least confirms the relative timing. The <a href="https://www.wunderground.com/history/daily/KIGX/date/2014-3-19?req_city=Durham&req_state=NC&req_statename=North%20Carolina&reqdb.zip=27708&reqdb.magic=1&reqdb.wmo=99999">precipitous weather</a> on March 14, 2014 in Durham, North Carolina supports, but does not confirm, that this day is a potential capture date.</p> +>>>>>>> 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 </section><section><div class='columns columns-2'><div class='column'><table> <thead><tr> <th>Camera</th> @@ -338,6 +345,7 @@ </tr> </tbody> </table> +<<<<<<< HEAD </div></div></section><section><h4>Errata</h4> <ul> <li>The Duke MTMC dataset paper mentions 2,700 identities, but their ground truth file only lists annotations for 1,812.</li> @@ -352,6 +360,13 @@ year = {2016} } </pre></section><section> +======= +</div></div></section><section><h4>Notes</h4> +<p>The original Duke MTMC dataset paper mentions 2,700 identities, but their ground truth file only lists annotations for 1,812, and their own research typically mentions 2,000. For this write up we used 2,000 to describe the approximate number of students.</p> +<h4>Ethics</h4> +<p>Please direct any questions about the ethics of the dataset to Duke University's <a href="https://hr.duke.edu/policies/expectations/compliance/">Institutional Ethics & Compliance Office</a> using the number at the bottom of the page.</p> +</section><section> +>>>>>>> 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 <h4>Cite Our Work</h4> <p> @@ -368,6 +383,7 @@ }</pre> </p> +<<<<<<< HEAD </section><section><h4>ToDo</h4> <ul> <li>clean up citations, formatting</li> @@ -376,6 +392,20 @@ </li><li>2 <a name="[^sensetime_qz]" class="footnote_shim"></a><span class="backlinks"><a href="#[^sensetime_qz]_1">a</a></span><a href="https://qz.com/1248493/sensetime-the-billion-dollar-alibaba-backed-ai-company-thats-quietly-watching-everyone-in-china/">https://qz.com/1248493/sensetime-the-billion-dollar-alibaba-backed-ai-company-thats-quietly-watching-everyone-in-china/</a> </li><li>3 <a name="[^sensenets_uyghurs]" class="footnote_shim"></a><span class="backlinks"><a href="#[^sensenets_uyghurs]_1">a</a></span><a href="https://foreignpolicy.com/2019/03/19/962492-orwell-china-socialcredit-surveillance/">https://foreignpolicy.com/2019/03/19/962492-orwell-china-socialcredit-surveillance/</a> </li><li>4 <a name="[^xinjiang_nyt]" class="footnote_shim"></a><span class="backlinks"><a href="#[^xinjiang_nyt]_1">a</a></span>Mozur, Paul. "One Month, 500,000 Face Scans: How China Is Using A.I. to Profile a Minority". <a href="https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html">https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html</a>. April 14, 2019. +======= +</section><section><p>If you use any data from the Duke MTMC please follow their <a href="http://vision.cs.duke.edu/DukeMTMC/#how-to-cite">license</a> and cite their work as:</p> +<pre> +@inproceedings{ristani2016MTMC, + title = {Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking}, + author = {Ristani, Ergys and Solera, Francesco and Zou, Roger and Cucchiara, Rita and Tomasi, Carlo}, + booktitle = {European Conference on Computer Vision workshop on Benchmarking Multi-Target Tracking}, + year = {2016} +} +</pre></section><section><h3>References</h3><section><ul class="footnotes"><li><a name="[^xinjiang_nyt]" class="footnote_shim"></a><span class="backlinks"><a href="#[^xinjiang_nyt]_1">a</a></span><p>Mozur, Paul. "One Month, 500,000 Face Scans: How China Is Using A.I. to Profile a Minority". <a href="https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html">https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html</a>. April 14, 2019.</p> +</li><li><a name="[^sensetime_qz]" class="footnote_shim"></a><span class="backlinks"><a href="#[^sensetime_qz]_1">a</a></span><p><a href="https://qz.com/1248493/sensetime-the-billion-dollar-alibaba-backed-ai-company-thats-quietly-watching-everyone-in-china/">https://qz.com/1248493/sensetime-the-billion-dollar-alibaba-backed-ai-company-thats-quietly-watching-everyone-in-china/</a></p> +</li><li><a name="[^sensenets_uyghurs]" class="footnote_shim"></a><span class="backlinks"><a href="#[^sensenets_uyghurs]_1">a</a></span><p><a href="https://foreignpolicy.com/2019/03/19/962492-orwell-china-socialcredit-surveillance/">https://foreignpolicy.com/2019/03/19/962492-orwell-china-socialcredit-surveillance/</a></p> +</li><li><a name="[^duke_mtmc_orig]" class="footnote_shim"></a><span class="backlinks"><a href="#[^duke_mtmc_orig]_1">a</a></span><p>"Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking". 2016. <a href="https://www.semanticscholar.org/paper/Performance-Measures-and-a-Data-Set-for-Tracking-Ristani-Solera/27a2fad58dd8727e280f97036e0d2bc55ef5424c">SemanticScholar</a></p> +>>>>>>> 61fbcb8f2709236f36a103a73e0bd9d1dd3723e8 </li></ul></section></section> </div> diff --git a/site/public/datasets/msceleb/index.html b/site/public/datasets/msceleb/index.html index 84c62bd2..86741647 100644 --- a/site/public/datasets/msceleb/index.html +++ b/site/public/datasets/msceleb/index.html @@ -49,8 +49,7 @@ </div><div class='meta'> <div class='gray'>Website</div> <div><a href='http://www.msceleb.org/' target='_blank' rel='nofollow noopener'>msceleb.org</a></div> - </div></div><p>[ PAGE UNDER DEVELOPMENT ]</p> -<p><a href="https://www.hrw.org/news/2019/01/15/letter-microsoft-face-surveillance-technology">https://www.hrw.org/news/2019/01/15/letter-microsoft-face-surveillance-technology</a></p> + </div></div><p><a href="https://www.hrw.org/news/2019/01/15/letter-microsoft-face-surveillance-technology">https://www.hrw.org/news/2019/01/15/letter-microsoft-face-surveillance-technology</a></p> <p><a href="https://www.scmp.com/tech/science-research/article/3005733/what-you-need-know-about-sensenets-facial-recognition-firm">https://www.scmp.com/tech/science-research/article/3005733/what-you-need-know-about-sensenets-facial-recognition-firm</a></p> </section><section> <h3>Who used Microsoft Celeb?</h3> diff --git a/site/public/datasets/oxford_town_centre/index.html b/site/public/datasets/oxford_town_centre/index.html index 2c7c26fc..03d8934b 100644 --- a/site/public/datasets/oxford_town_centre/index.html +++ b/site/public/datasets/oxford_town_centre/index.html @@ -128,7 +128,7 @@ title = {MegaPixels: Origins, Ethics, and Privacy Implications of Publicly Available Face Recognition Image Datasets}, year = 2019, url = {https://megapixels.cc/}, - urldate = {2019-04-20} + urldate = {2019-04-18} }</pre> </p> diff --git a/site/public/datasets/uccs/index.html b/site/public/datasets/uccs/index.html index 9347d536..3652e329 100644 --- a/site/public/datasets/uccs/index.html +++ b/site/public/datasets/uccs/index.html @@ -49,14 +49,14 @@ </div><div class='meta'> <div class='gray'>Website</div> <div><a href='http://vast.uccs.edu/Opensetface/' target='_blank' rel='nofollow noopener'>uccs.edu</a></div> - </div></div><p>UnConstrained College Students (UCCS) is a dataset of long-range surveillance photos captured at University of Colorado Colorado Springs developed primarily for research and development of "face detection and recognition research towards surveillance applications"<a class="footnote_shim" name="[^uccs_vast]_1"> </a><a href="#[^uccs_vast]" class="footnote" title="Footnote 1">1</a>. According to the authors of two papers associated with the dataset, over 1,700 students and pedestrians were "photographed using a long-range high-resolution surveillance camera without their knowledge".<a class="footnote_shim" name="[^funding_uccs]_1"> </a><a href="#[^funding_uccs]" class="footnote" title="Footnote 3">3</a> In this investigation, we examine the contents of the dataset, funding sources, photo EXIF data, and information from publicly available research project citations.</p> + </div></div><p>UnConstrained College Students (UCCS) is a dataset of long-range surveillance photos captured at University of Colorado Colorado Springs developed primarily for research and development of "face detection and recognition research towards surveillance applications"<a class="footnote_shim" name="[^uccs_vast]_1"> </a><a href="#[^uccs_vast]" class="footnote" title="Footnote 1">1</a>. According to the authors of <a href="https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745">two</a> <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1">papers</a> associated with the dataset, over 1,700 students and pedestrians were "photographed using a long-range high-resolution surveillance camera without their knowledge".<a class="footnote_shim" name="[^funding_uccs]_1"> </a><a href="#[^funding_uccs]" class="footnote" title="Footnote 3">3</a> In this investigation, we examine the contents of the <a href="http://vast.uccs.edu/Opensetface/">dataset</a>, its funding sources, photo EXIF data, and information from publicly available research project citations.</p> <p>The UCCS dataset includes over 1,700 unique identities, most of which are students walking to and from class. As of 2018, it was the "largest surveillance [face recognition] benchmark in the public domain."<a class="footnote_shim" name="[^surv_face_qmul]_1"> </a><a href="#[^surv_face_qmul]" class="footnote" title="Footnote 4">4</a> The photos were taken during the spring semesters of 2012 – 2013 on the West Lawn of the University of Colorado Colorado Springs campus. The photographs were timed to capture students during breaks between their scheduled classes in the morning and afternoon during Monday through Thursday. "For example, a student taking Monday-Wednesday classes at 12:30 PM will show up in the camera on almost every Monday and Wednesday."<a class="footnote_shim" name="[^sapkota_boult]_1"> </a><a href="#[^sapkota_boult]" class="footnote" title="Footnote 2">2</a>.</p> -</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_grid.jpg' alt=' Example images from the UnConstrained College Students Dataset. '><div class='caption'> Example images from the UnConstrained College Students Dataset. </div></div></section><section><p>The long-range surveillance images in the UnContsrained College Students dataset were taken using a Canon 7D 18-megapixel digital camera fitted with a Sigma 800mm F5.6 EX APO DG HSM telephoto lens and pointed out an office window across the university's West Lawn. The students were photographed from a distance of approximately 150 meters through an office window. "The camera [was] programmed to start capturing images at specific time intervals between classes to maximize the number of faces being captured."<a class="footnote_shim" name="[^sapkota_boult]_2"> </a><a href="#[^sapkota_boult]" class="footnote" title="Footnote 2">2</a> -Their setup made it impossible for students to know they were being photographed, providing the researchers with realistic surveillance images to help build face recognition systems for real world applications in defense, intelligence, and commercial sectors.</p> -</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_map_aerial.jpg' alt=' The location at University of Colorado Colorado Springs where students were surreptitiously photographed with a long-range surveillance camera for use in a defense and intelligence agency funded research project on face recognition. Image: Google Maps'><div class='caption'> The location at University of Colorado Colorado Springs where students were surreptitiously photographed with a long-range surveillance camera for use in a defense and intelligence agency funded research project on face recognition. Image: Google Maps</div></div></section><section><p>The EXIF data embedded in the images shows that the photo capture times follow a similar pattern, but also highlights that the vast majority of photos (over 7,000) were taken on Tuesdays around noon during students' lunch break. The lack of any photos taken on Friday shows that the researchers were only interested in capturing images of students.</p> -</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_exif_plot_days.png' alt=' UCCS photos captured per weekday © megapixels.cc'><div class='caption'> UCCS photos captured per weekday © megapixels.cc</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_exif_plot.png' alt=' UCCS photos captured per weekday © megapixels.cc'><div class='caption'> UCCS photos captured per weekday © megapixels.cc</div></div></section><section><p>The two research papers associated with the release of the UCCS dataset (<a href="https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745">Unconstrained Face Detection and Open-Set Face Recognition Challenge</a> and <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1">Large Scale Unconstrained Open Set Face Database</a>), acknowledge that the primary funding sources for their work were United States defense and intelligence agencies. Specifically, development of the UnContsrianed College Students dataset was funded by the Intelligence Advanced Research Projects Activity (IARPA), Office of Director of National Intelligence (ODNI), Office of Naval Research and The Department of Defense Multidisciplinary University Research Initiative (ONR MURI), and the Special Operations Command and Small Business Innovation Research (SOCOM SBIR) amongst others. UCCS's VAST site also explicitly <a href="https://vast.uccs.edu/project/iarpa-janus/">states</a> that they are part of the <a href="https://www.iarpa.gov/index.php/research-programs/janus">IARPA Janus</a>, a face recognition project developed to serve the needs of national intelligence interests, clearly establishing the the funding sources and immediate benefactors of this dataset are United States defense and intelligence agencies.</p> -<p>Although the images were first captured in 2012 – 2013 the dataset was not publicly released until 2016. In 2017 the UCCS face dataset formed the basis for a defense and intelligence agency funded <a href="http://www.face-recognition-challenge.com/">face recognition challenge</a> project at the International Joint Biometrics Conference in Denver, CO. And in 2018 the dataset was again used for the <a href="https://erodner.github.io/ial2018eccv/">2nd Unconstrained Face Detection and Open Set Recognition Challenge</a> at the European Computer Vision Conference (ECCV) in Munich, Germany.</p> -<p>As of April 15, 2019, the UCCS dataset is no longer available for public download. But during the three years it was publicly available (2016-2019) the UCCS dataset appeared in at least 6 publicly available research papers including verified usage from Beihang University who is known to provide research and development for China's military.</p> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_map_aerial.jpg' alt=' The location at University of Colorado Colorado Springs where students were surreptitiously photographed with a long-range surveillance camera for use in a defense and intelligence agency funded research project on face recognition. Image: Google Maps'><div class='caption'> The location at University of Colorado Colorado Springs where students were surreptitiously photographed with a long-range surveillance camera for use in a defense and intelligence agency funded research project on face recognition. Image: Google Maps</div></div></section><section><p>The long-range surveillance images in the UnContsrained College Students dataset were taken using a Canon 7D 18-megapixel digital camera fitted with a Sigma 800mm F5.6 EX APO DG HSM telephoto lens and pointed out an office window across the university's West Lawn. The students were photographed from a distance of approximately 150 meters through an office window. "The camera [was] programmed to start capturing images at specific time intervals between classes to maximize the number of faces being captured."<a class="footnote_shim" name="[^sapkota_boult]_2"> </a><a href="#[^sapkota_boult]" class="footnote" title="Footnote 2">2</a> +Their setup made it impossible for students to know they were being photographed, providing the researchers with realistic surveillance images to help build face recognition systems for real world applications for defense, intelligence, and commercial partners.</p> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_grid.jpg' alt=' Example images from the UnConstrained College Students Dataset. '><div class='caption'> Example images from the UnConstrained College Students Dataset. </div></div></section><section><p>The EXIF data embedded in the images shows that the photo capture times follow a similar pattern to that outlined by the researchers, but also highlights that the vast majority of photos (over 7,000) were taken on Tuesdays around noon during students' lunch break. The lack of any photos taken between Friday through Sunday shows that the researchers were only interested in capturing images of students during the peak campus hours.</p> +</section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_exif_plot_days.png' alt=' UCCS photos captured per weekday © megapixels.cc'><div class='caption'> UCCS photos captured per weekday © megapixels.cc</div></div></section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_exif_plot.png' alt=' UCCS photos captured per weekday © megapixels.cc'><div class='caption'> UCCS photos captured per weekday © megapixels.cc</div></div></section><section><p>The two research papers associated with the release of the UCCS dataset (<a href="https://www.semanticscholar.org/paper/Unconstrained-Face-Detection-and-Open-Set-Face-G%C3%BCnther-Hu/d4f1eb008eb80595bcfdac368e23ae9754e1e745">Unconstrained Face Detection and Open-Set Face Recognition Challenge</a> and <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1">Large Scale Unconstrained Open Set Face Database</a>), acknowledge that the primary funding sources for their work were United States defense and intelligence agencies. Specifically, development of the UnContsrianed College Students dataset was funded by the Intelligence Advanced Research Projects Activity (IARPA), Office of Director of National Intelligence (ODNI), Office of Naval Research and The Department of Defense Multidisciplinary University Research Initiative (ONR MURI), and the Special Operations Command and Small Business Innovation Research (SOCOM SBIR) amongst others. UCCS's VAST site also explicitly <a href="https://vast.uccs.edu/project/iarpa-janus/">states</a> their involvement in the <a href="https://www.iarpa.gov/index.php/research-programs/janus">IARPA Janus</a> face recognition project developed to serve the needs of national intelligence, establishing that immediate benefactors of this dataset include United States defense and intelligence agencies, but it would go on to benefit other similar organizations.</p> +<p>In 2017, one year after its public release, the UCCS face dataset formed the basis for a defense and intelligence agency funded <a href="http://www.face-recognition-challenge.com/">face recognition challenge</a> project at the International Joint Biometrics Conference in Denver, CO. And in 2018 the dataset was again used for the <a href="https://erodner.github.io/ial2018eccv/">2nd Unconstrained Face Detection and Open Set Recognition Challenge</a> at the European Computer Vision Conference (ECCV) in Munich, Germany.</p> +<p>As of April 15, 2019, the UCCS dataset is no longer available for public download. But during the three years it was publicly available (2016-2019) the UCCS dataset appeared in at least 6 publicly available research papers including verified usage from Beihang University who is known to provide research and development for China's military; and Vision Semantics Ltd who lists the UK Ministory of Defence as a project partner.</p> </section><section> <h3>Who used UCCS?</h3> @@ -117,7 +117,7 @@ Their setup made it impossible for students to know they were being photographed <h2>Supplementary Information</h2> -</section><section><p>To show the types of face images used in the UCCS student dataset while protecting their individual privacy, a generative adversarial network was used to interpolate between identities in the dataset. The image below shows a generative adversarial network trained on the UCCS face bounding box areas from 16,000 images and over 90,000 face regions.</p> +</section><section><p>Since this site To show the types of face images used in the UCCS student dataset while protecting their individual privacy, a generative adversarial network was used to interpolate between identities in the dataset. The image below shows a generative adversarial network trained on the UCCS face bounding box areas from 16,000 images and over 90,000 face regions.</p> </section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_pgan_01.jpg' alt=' GAN generated approximations of students in the UCCS dataset. © megapixels.cc 2018'><div class='caption'> GAN generated approximations of students in the UCCS dataset. © megapixels.cc 2018</div></div></section><section><div class='columns columns-2'><div class='column'><h4>UCCS photos taken in 2012</h4> <table> <thead><tr> @@ -211,7 +211,7 @@ Their setup made it impossible for students to know they were being photographed </tbody> </table> </div></div></section><section><h3>Location</h3> -<p>The location of the camera and subjects can confirmed using several visual cues in the dataset images: the unique pattern of the sidewalk that is only used on the UCCS Pedestrian Spine near the West Lawn, the two UCCS sign poles with matching graphics still visible in Google Street View, the no parking sign and directionality of its arrow, the back of street sign next to it, the slight bend in the sidewalk, the presence of cars passing in the background of the image, and the far wall of the parking garage all match images in the dataset. The <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1">original papers</a> also provides another clue: a <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1/figure/1">picture of the camera</a> inside the office that was used to create the dataset. The window view in this image provides another match for the brick pattern on the north facade of the Kraember Family Library and the green metal fence along the sidewalk. View the <a href="https://www.google.com/maps/place/University+of+Colorado+Colorado+Springs/@38.8934297,-104.7992445,27a,35y,258.51h,75.06t/data=!3m1!1e3!4m5!3m4!1s0x87134fa088fe399d:0x92cadf3962c058c4!8m2!3d38.8968312!4d-104.8049528">location on Google Maps</a></p> +<p>The location of the camera and subjects was confirmed using several visual cues in the dataset images: the unique pattern of the sidewalk that is only used on the UCCS Pedestrian Spine near the West Lawn, the two UCCS sign poles with matching graphics still visible in Google Street View, the no parking sign and directionality of its arrow, the back of street sign next to it, the slight bend in the sidewalk, the presence of cars passing in the background of the image, and the far wall of the parking garage all match images in the dataset. The <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1">original papers</a> also provides another clue: a <a href="https://www.semanticscholar.org/paper/Large-scale-unconstrained-open-set-face-database-Sapkota-Boult/07fcbae86f7a3ad3ea1cf95178459ee9eaf77cb1/figure/1">picture of the camera</a> inside the office that was used to create the dataset. The window view in this image provides another match for the brick pattern on the north facade of the Kraember Family Library and the green metal fence along the sidewalk. View the <a href="https://www.google.com/maps/place/University+of+Colorado+Colorado+Springs/@38.8934297,-104.7992445,27a,35y,258.51h,75.06t/data=!3m1!1e3!4m5!3m4!1s0x87134fa088fe399d:0x92cadf3962c058c4!8m2!3d38.8968312!4d-104.8049528">location on Google Maps</a></p> </section><section class='images'><div class='image'><img src='https://nyc3.digitaloceanspaces.com/megapixels/v1/datasets/uccs/assets/uccs_map_3d.jpg' alt=' 3D view showing the angle of view of the surveillance camera used for UCCS dataset. Image: Google Maps'><div class='caption'> 3D view showing the angle of view of the surveillance camera used for UCCS dataset. Image: Google Maps</div></div></section><section><h3>Funding</h3> <p>The UnConstrained College Students dataset is associated with two main research papers: "Large Scale Unconstrained Open Set Face Database" and "Unconstrained Face Detection and Open-Set Face Recognition Challenge". Collectively, these papers and the creation of the dataset have received funding from the following organizations:</p> <ul> @@ -246,7 +246,7 @@ Their setup made it impossible for students to know they were being photographed title = {MegaPixels: Origins, Ethics, and Privacy Implications of Publicly Available Face Recognition Image Datasets}, year = 2019, url = {https://megapixels.cc/}, - urldate = {2019-04-20} + urldate = {2019-04-18} }</pre> </p> |
