Merge branch 'master' of github.com:adamhrv/megapixels_dev

1 files changed, 262 insertions, 0 deletions

diff --git a/megapixels/app/server/tasks/demo.py b/megapixels/app/server/tasks/demo.py
new file mode 100644
index 00000000..c27b08b5
--- /dev/null
+++ b/megapixels/app/server/tasks/demo.py
@@ -0,0 +1,262 @@
+
+import app.settings.app_cfg as cfg
+from app.server.tasks import celery
+
+from celery.utils.log import get_task_logger
+log = get_task_logger(__name__)
+
+opt_size = (256, 256,)
+
+@celery.task(bind=True)
+def demo_task(self, uuid_name, fn):
+
+  import sys
+  import os
+  from os.path import join
+  from pathlib import Path
+  import time
+
+  import numpy as np
+  import cv2 as cv
+  import dlib
+  from PIL import Image
+  import matplotlib.pyplot as plt
+
+  from app.utils import logger_utils, file_utils, im_utils, display_utils, draw_utils
+  from app.utils import plot_utils
+  from app.processors import face_detector, face_landmarks, face_age_gender, face_beauty
+  # , face_emotion
+  from app.models.data_store import DataStore
+
+  # TODO add selective testing
+  opt_gpu = -1
+  opt_run_pose = True
+  opt_run_2d_68 = True
+  opt_run_3d_68 = True
+  opt_run_3d_68 = True
+
+  opt_gif_size = (256, 256,)
+  opt_gif_frames = 15
+
+  meta = {
+    'step': 0,
+    'total': 10,
+    'message': 'Starting',
+    'uuid': uuid_name,
+    'data': { 'statistics': {} },
+  }
+  paths = []
+
+  def step(msg, step=1):
+    meta['message'] = msg
+    meta['step'] += step
+    log.debug('> {}'.format(msg))
+    self.update_state(state='PROCESSING', meta=meta)
+
+  def save_image(key, title, data):
+    fn =  '{}_{}.jpg'.format(uuid_name, key)
+    fpath = os.path.join(cfg.DIR_SITE_USER_CONTENT, fn)
+    paths.append(fpath)
+    cv.imwrite(fpath, data)
+
+    meta['data'][key] = {
+      'title': title,
+      'url': os.path.join('/user_content/', fn),
+    }
+
+  step('Loading image')
+  self.update_state(state='PROCESSING', meta=meta)
+
+  # os.path.join('/user_content/', fn)
+
+  # -------------------------------------------------
+  # init here
+
+  # load image
+  im = cv.imread(fn)
+  im_resized = im_utils.resize(im, width=opt_size[0], height=opt_size[1])
+
+  # ----------------------------------------------------------------------------
+  # detect face
+
+  face_detector_instance = face_detector.DetectorDLIBCNN(gpu=opt_gpu)  # -1 for CPU
+  step('Detecting face')
+  st = time.time()
+  bboxes = face_detector_instance.detect(im_resized, largest=True)
+  bbox = bboxes[0]
+  dim = im_resized.shape[:2][::-1]
+  bbox_dim = bbox.to_dim(dim)
+  if not bbox:
+    log.error('No face detected')
+    meta['error'] = 'No face detected'
+    self.update_state(state='FAILURE', meta=meta)
+    return meta
+  else:
+    log.info(f'Detected face in {(time.time() - st):.2f}s')
+
+
+  # ----------------------------------------------------------------------------
+  # detect 3D landmarks
+  
+  step('Generating 3D Landmarks')
+  log.info('loading 3D landmark generator files...')
+  landmark_detector_3d_68 = face_landmarks.FaceAlignment3D_68(gpu=opt_gpu)  # -1 for CPU
+  log.info('generating 3D landmarks...')
+  st = time.time()
+  points_3d_68 = landmark_detector_3d_68.landmarks(im_resized, bbox_dim.to_xyxy())
+  log.info(f'generated 3D landmarks in {(time.time() - st):.2f}s')
+  log.info('')
+
+  # draw 3d landmarks
+  im_landmarks_3d_68 = im_resized.copy()
+  draw_utils.draw_landmarks3D(im_landmarks_3d_68, points_3d_68)
+  draw_utils.draw_bbox(im_landmarks_3d_68, bbox_dim)
+
+  save_image('landmarks_3d_68', '3D Landmarks', im_landmarks_3d_68)
+
+  # ----------------------------------------------------------------------------
+  # generate 3D GIF animation
+  
+  step('Generating GIF Animation')
+  log.info('generating 3D animation...')
+
+  fn =  '{}_{}.gif'.format(uuid_name, '3d')
+  fp_out = os.path.join(cfg.DIR_SITE_USER_CONTENT, fn)
+  paths.append(fp_out)
+
+  st = time.time()
+  plot_utils.generate_3d_landmark_anim(np.array(points_3d_68), fp_out, 
+    size=opt_gif_size, num_frames=opt_gif_frames)
+  log.info(f'Generated animation in {(time.time() - st):.2f}s')
+  log.info(f'Saved to: {fp_out}')
+  log.info('')
+
+  meta['data']['points_3d_68'] = points_3d_68
+  meta['data']['points_3d_68'] = {
+    'title': '3D Animated GIF',
+    'url': os.path.join('/user_content/', fn),
+  }
+
+  # ----------------------------------------------------------------------------
+  # generate 68 point landmarks using dlib
+  
+  step('Generating 2D 68PT landmarks')
+  log.info('initializing face landmarks 68 dlib...')
+  from app.processors import face_landmarks
+  landmark_detector_2d_68 = face_landmarks.Dlib2D_68()
+  log.info('generating 2D 68PT landmarks...')
+  st = time.time()
+  points_2d_68 = landmark_detector_2d_68.landmarks(im_resized, bbox_dim)
+  log.info(f'generated 2D 68PT face landmarks in {(time.time() - st):.2f}s')
+  log.info('')
+
+  # draw 2d landmarks
+  im_landmarks_2d_68 = im_resized.copy()
+  draw_utils.draw_landmarks2D(im_landmarks_2d_68, points_2d_68)
+  draw_utils.draw_bbox(im_landmarks_2d_68, bbox_dim)
+  save_image('landmarks_2d_68', '2D Landmarks', im_landmarks_2d_68)
+
+  # ----------------------------------------------------------------------------
+  # generate pose from 68 point 2D landmarks
+  
+  if opt_run_pose:
+    step('Generating pose')
+    log.info('initialize pose...')
+    from app.processors import face_pose
+    pose_detector = face_pose.FacePoseDLIB()
+    log.info('generating pose...')
+    st = time.time()
+    pose_data = pose_detector.pose(points_2d_68, dim)
+    log.info(f'generated pose {(time.time() - st):.2f}s')
+    log.info('')
+
+    im_pose = im_resized.copy()
+    draw_utils.draw_pose(im_pose, pose_data['point_nose'], pose_data['points'])
+    draw_utils.draw_degrees(im_pose, pose_data)
+    save_image('pose', 'Pose', im_pose)
+
+  # ----------------------------------------------------------------------------
+  # age
+
+  # real
+  step('Running age predictor')
+  age_real_predictor = face_age_gender.FaceAgeReal()
+  st = time.time()
+  age_real = age_real_predictor.predict(im_resized, bbox_dim)
+  log.info(f'age real took: {(time.time()-st)/1000:.5f}s')
+  meta['data']['statistics']['age_real'] = f'{(age_real):.2f}'
+  
+  # apparent
+  age_apparent_predictor = face_age_gender.FaceAgeApparent()
+  st = time.time()
+  age_apparent = age_apparent_predictor.predict(im_resized, bbox_dim)
+  log.info(f'age apparent took: {(time.time()-st)/1000:.5f}s')
+  meta['data']['statistics']['age_apparent'] = f'{(age_apparent):.2f}'
+
+  # gender
+  step('Running gender predictor')
+  gender_predictor = face_age_gender.FaceGender()
+  st = time.time()
+  gender = gender_predictor.predict(im_resized, bbox_dim)
+  log.info(f'gender took: {(time.time()-st)/1000:.5f}s')
+  meta['data']['statistics']['gender'] = f"M: {gender['m']:.2f}, F: {gender['f']:.2f}"
+
+  # # ----------------------------------------------------------------------------
+  # # emotion
+
+  # emotion_predictor = face_emotion.FaceEmotion(gpu=opt_gpu)
+  # emotion_score = emotion_predictor.emotion(im_resized, bbox_dim)
+  # log.info(f'emotion score: {(100*emotion_score):.2f}')
+
+  # im_emotion = im_resized.copy()
+  # draw_utils.draw_bbox(im_emotion, bbox_dim)
+  # txt = f'emotion score: {(100*emotion_score):.2f}'
+  # draw_utils.draw_text(im_emotion, bbox_dim.pt_tl, txt)
+  # save_image('emotion', 'Emotion', im_emotion)
+
+
+  # ----------------------------------------------------------------------------
+  # beauty
+
+  # TODO fix Keras CPU/GPU device selection issue
+  # NB: GPU visibility issues with dlib/keras
+  # Wrap this with cuda toggle and run before init dlib GPU
+
+  step('Running beauty predictor')
+  device_cur = os.getenv('CUDA_VISIBLE_DEVICES', '')
+  os.environ['CUDA_VISIBLE_DEVICES'] = ''
+  beauty_predictor = face_beauty.FaceBeauty()
+  os.environ['CUDA_VISIBLE_DEVICES'] = device_cur
+
+  beauty_score = beauty_predictor.beauty(im_resized, bbox_dim)
+  log.info(f'beauty score: {(100*beauty_score):.2f}')
+
+  # # draw 2d landmarks
+  # im_beauty = im_resized.copy()
+  # draw_utils.draw_bbox(im_beauty, bbox_dim)
+  # txt = f'Beauty score: {(100*beauty_score):.2f}'
+  # draw_utils.draw_text(im_beauty, bbox_dim.pt_tl, txt)
+  # save_image('beauty', 'Beauty', im_beauty)
+  meta['data']['statistics']['beauty'] = f'{(100*beauty_score):.2f}'
+
+  step('Done')
+  
+  # # 3DDFA
+  # self.log.debug('Add depth')
+  # self.log.debug('Add pncc')
+
+  # # TODO
+  # self.log.debug('Add 3D face model')
+  # self.log.debug('Add face texture flat')
+  # self.log.debug('Add ethnicity')
+
+  log.debug('done!!')
+
+  time.sleep(3)
+  for path in paths:
+    if os.path.exists(path):
+      os.remove(path)
+  
+  meta['step'] = meta['total']
+  meta['state'] = 'SUCCESS'
+  return meta