diff options
| author | Jules Laplace <julescarbon@gmail.com> | 2020-01-08 01:54:28 +0100 |
|---|---|---|
| committer | Jules Laplace <julescarbon@gmail.com> | 2020-01-08 01:54:28 +0100 |
| commit | a194eaa66108d753aac1eac70b7016a9b20897e1 (patch) | |
| tree | 2e8678c9e0b5b3c6a1aa5587ad087c66dbfb85f0 /cli/app | |
| parent | daa10de26027528528f485b6dfed94864faf259a (diff) | |
up
Diffstat (limited to 'cli/app')
| -rw-r--r-- | cli/app/commands/biggan/search_class.py | 4 | ||||
| -rw-r--r-- | cli/app/search/search_class.py | 2 |
2 files changed, 3 insertions, 3 deletions
diff --git a/cli/app/commands/biggan/search_class.py b/cli/app/commands/biggan/search_class.py index 066062f..6e1df95 100644 --- a/cli/app/commands/biggan/search_class.py +++ b/cli/app/commands/biggan/search_class.py @@ -20,9 +20,9 @@ from app.utils.cortex_utils import cortex_folder, download_cortex_files, find_un @click.option('-t', '--tag', 'opt_tag', default='inverse_' + str(int(time.time() * 1000)), help='Tag this dataset') @click.option('-sc', '--stochastic_clipping', 'opt_stochastic_clipping', default=0, - help='Compute feature loss') + help='Compute feature loss. Specify interval to do the clipping') @click.option('-lc', '--label_clipping', 'opt_label_clipping', default=0, - help='Normalize labels every N steps') + help='Normalize labels every N steps. Specify interval to do the normalization') @click.option('-feat', '--use_feature_detector', 'opt_use_feature_detector', is_flag=True, help='Compute feature loss') @click.option('-ll', '--feature_layers', 'opt_feature_layers', default="1a,2a,4a,7a", diff --git a/cli/app/search/search_class.py b/cli/app/search/search_class.py index a4bec07..cd53a71 100644 --- a/cli/app/search/search_class.py +++ b/cli/app/search/search_class.py @@ -116,7 +116,7 @@ def find_nearest_vector(sess, generator, opt_fp_in, opt_dims, out_images, out_la ## if computing Feature loss, use these encoders if opt_use_feature_detector: print("Initializing feature detector...") - pix_square_diff = tf.square((target_img - gen_img) / 2.0) + pix_square_diff = tf.square((target - output) / 2.0) mse_loss = tf.reduce_mean(pix_square_diff) feature_extractor = hub.Module("https://tfhub.dev/google/imagenet/inception_v3/feature_vector/1") |
