def check_args(args):
'''determine if is None'''
assert (
(args.rpn_pkl is not None and args.rpn_cfg is not None) or
(args.rpn_pkl is None and args.rpn_cfg is None)
)
if args.rpn_pkl is not None:
'''Download the file specified by the URL to the cache_dir and return the path to the cached file.
If the argument is not a URL, simply return it as is.'''
args.rpn_pkl = cache_url(args.rpn_pkl, cfg.DOWNLOAD_CACHE)
assert os.path.exists(args.rpn_pkl)
assert os.path.exists(args.rpn_cfg)
if args.models_to_run is not None:
'''determine if length ==2'''
assert len(args.models_to_run) % 2 == 0
for i, model_file in enumerate(args.models_to_run):
if len(model_file) > 0:
if i % 2 == 0:
'''download 2nd model file'''
model_file = cache_url(model_file, cfg.DOWNLOAD_CACHE)
args.models_to_run[i] = model_file
assert os.path.exists(model_file), '\'{}\' does not exist'.format(model_file)
def cache_cfg_urls():
"""Download URLs in the config, cache them locally, and rewrite cfg to make
use of the locally cached file.
"""
__C.TRAIN.WEIGHTS = cache_url(__C.TRAIN.WEIGHTS, __C.DOWNLOAD_CACHE)
__C.TEST.WEIGHTS = cache_url(__C.TEST.WEIGHTS, __C.DOWNLOAD_CACHE)
__C.TRAIN.PROPOSAL_FILES = tuple(
[cache_url(f, __C.DOWNLOAD_CACHE) for f in __C.TRAIN.PROPOSAL_FILES]
)
__C.TEST.PROPOSAL_FILES = tuple(
[cache_url(f, __C.DOWNLOAD_CACHE) for f in __C.TEST.PROPOSAL_FILES]
)
def main(args):
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
cap = cv2.VideoCapture(0)
while True:
ret, frame = cap.read()
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, timers=timers)
image = vis_utils.vis_one_image_opencv(np.array(frame),
cls_boxes,
cls_segms,
cls_keyps,
thresh=0.7,
kp_thresh=2,
show_box=True,
show_class=True)
cv2.imshow('camera', image)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
print("Time:", time.time() - t)
cap.release()
cv2.destroyAllWindows()
def build_graph(self):
c2_utils.import_detectron_ops()
# OpenCL may be enabled by default in OpenCV3; disable it because it's not
# thread safe and causes unwanted GPU memory allocations.
cv2.ocl.setUseOpenCL(False)
merge_cfg_from_file(self.config.args['config_path'])
# If this is a CPU kernel, tell Caffe2 that it should not use
# any GPUs for its graph operations
cpu_only = True
for handle in self.config.devices:
if handle.type == DeviceType.GPU.value:
cpu_only = False
if cpu_only:
cfg.NUM_GPUS = 0
else:
cfg.NUM_GPUS = 1
# TODO: wrap this in "with device"
weights_path = cache_url(self.config.args['weights_path'],
cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(weights_path)
return model
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
start = timeit.default_timer()
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + "/*." + args.image_ext)
else:
im_list = [args.im_or_folder]
# extract bboxes from bottom-up attention model
image_bboxes = {}
if args.bbox_file is not None:
image_bboxes = extract_bboxes(args.bbox_file)
count = 0
if not os.path.exists(args.output_dir):
os.makedirs(args.output_dir)
for i, im_name in enumerate(im_list):
im_base_name = os.path.basename(im_name)
image_id = int(im_base_name.split(".")[0].split("_")[-1]) # for COCO
if image_id % args.total_group == args.group_id:
bbox = image_bboxes[image_id] if image_id in image_bboxes else None
im = cv2.imread(im_name)
if im is not None:
outfile = os.path.join(
args.output_dir, im_base_name.replace("jpg", "npy")
)
lock_folder = outfile.replace("npy", "lock")
if not os.path.exists(lock_folder) and os.path.exists(outfile):
continue
if not os.path.exists(lock_folder):
os.makedirs(lock_folder)
result = get_detections_from_im(
cfg,
model,
im,
image_id,
args.feat_name,
args.min_bboxes,
args.max_bboxes,
background=args.background,
bboxes=bbox,
)
np.save(outfile, result)
os.rmdir(lock_folder)
count += 1
if count % 100 == 0:
end = timeit.default_timer()
epoch_time = end - start
print("process {:d} images after {:.1f} s".format(count, epoch_time))
def main(args):
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
# get the weight path
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
#dummy_coco_dataset = dummy_datasets.get_coco_dataset()
det_dir = os.path.join(args.input_dir, 'det')
if not os.path.exists(det_dir):
os.makedirs(det_dir)
txt_file = os.path.join(det_dir, "det.txt")
fid = open(txt_file, 'w')
img_dir = os.path.join(args.input_dir, "img1")
img_list = os.listdir(img_dir)
img_list = sorted(img_list)
for i in range(len(img_list)):
print("processing: %d/%d" % (i + 1, len(img_list)))
img_name = img_list[i][:-4]
img_idx = int(img_name)
img_path = os.path.join(img_dir, img_list[i])
frame = cv2.imread(img_path)
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, None)
vis_utils.write_txt(fid, img_idx, cls_boxes)
fid.close()
def check_args(args):
assert ((args.rpn_pkl is not None and args.rpn_cfg is not None)
or (args.rpn_pkl is None and args.rpn_cfg is None))
if args.rpn_pkl is not None:
args.rpn_pkl = cache_url(args.rpn_pkl, cfg.DOWNLOAD_CACHE)
assert os.path.exists(args.rpn_pkl)
assert os.path.exists(args.rpn_cfg)
if args.models_to_run is not None:
assert len(args.models_to_run) % 2 == 0
for i, model_file in enumerate(args.models_to_run):
if len(model_file) > 0:
if i % 2 == 0:
model_file = cache_url(model_file, cfg.DOWNLOAD_CACHE)
args.models_to_run[i] = model_file
assert os.path.exists(model_file), \
'\'{}\' does not exist'.format(model_file)
def check_args(args):
assert (
(args.rpn_pkl is not None and args.rpn_cfg is not None) or
(args.rpn_pkl is None and args.rpn_cfg is None)
)
if args.rpn_pkl is not None:
args.rpn_pkl = cache_url(args.rpn_pkl, cfg.DOWNLOAD_CACHE)
assert os.path.exists(args.rpn_pkl)
assert os.path.exists(args.rpn_cfg)
if args.models_to_run is not None:
assert len(args.models_to_run) % 2 == 0
for i, model_file in enumerate(args.models_to_run):
if len(model_file) > 0:
if i % 2 == 0:
model_file = cache_url(model_file, cfg.DOWNLOAD_CACHE)
args.models_to_run[i] = model_file
assert os.path.exists(model_file), \
'\'{}\' does not exist'.format(model_file)
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
# im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
im_list = search(args.im_or_folder, args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
#for i, im_name in im_list:
#out_name = os.path.join(
# args.output_dir, '{}'.format(os.path.basename(im_name) + '.pdf')
#)
out_name = im_name.replace(args.im_or_folder, args.output_dir)
par_path = os.path.dirname(out_name)
if not os.path.exists(par_path):
os.makedirs(par_path)
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers
)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)'
)
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
par_path,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2,
ext='png'
)
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = (
dummy_datasets.get_vg3k_dataset()
if args.use_vg3k else dummy_datasets.get_coco_dataset())
if os.path.isdir(args.im_or_folder):
im_list = sorted(glob.iglob(args.im_or_folder + '/*.' + args.image_ext))
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
out_name = os.path.join(
args.output_dir, '{}'.format(os.path.basename(im_name) + '.npz')
)
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
if im is None:
logger.info('Unable to read image, skipping.')
continue
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers
)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)'
)
boxes, segms, classes = convert(cls_boxes, cls_segms)
classes = np.array(classes, dtype=np.uint16)
resolution = segms[0]['size']
segms = np.array([x['counts'] for x in segms]) # Run-length encoding
valid = boxes[:, 4] >= args.thresh
if args.filter_classes:
valid &= np.isin(classes, all_classes)
boxes = boxes[valid].copy()
classes = classes[valid].copy()
segms = segms[valid].copy()
output_name = os.path.basename(im_name)
np.savez(args.output_dir + '/' + output_name, boxes=boxes, segments=segms, classes=classes, resolution=resolution)
def main(args):
logger = logging.getLogger(__name__)
args.cfg = '/detectron/configs/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_2x.yaml'
args.weights = '/data/workspace/fbdet/models/mask_rcnn_R_101_FPN_2x/model_final.pkl'
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
im_list = []
im_list.append('/data/workspace/fbdet/test_pic/11.jpg')
video = cv2.VideoCapture('/data/pic/valid_video_00.avi')
frame = 0
while (True):
if frame > 0:
break
# ret, im = video.read()
# if im is None or ret is None:
# print("video.read() fail || video.read() is end!")
# break
im = cv2.imread(im_list[0])
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
print('-----------------------------', frame)
boxs_list = vis_utils.get_boxes_image(cls_boxes,
cls_segms,
cls_keyps,
thresh=0.7,
dataset=dummy_coco_dataset)
print(boxs_list)
print('-----------------------------')
for i in range(len(boxs_list)):
box = boxs_list[i]
drawBoxOnImg(im, box[1], box[2], box[3], box[4], 0, 0, frame)
frame += 1
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
for root_dir_path_1, sub_dir_path_list_1, sub_file_path_list_1 in os.walk(
args.im_or_folder):
sub_dir_path_list_1 = sorted(sub_dir_path_list_1)
for i, sub_dir_path_1 in enumerate(sub_dir_path_list_1):
for root_dir_path_2, sub_dir_path_list_2, sub_file_path_list_2 in os.walk(
os.path.join(root_dir_path_1, sub_dir_path_1)):
sub_file_path_list_2 = sorted(sub_file_path_list_2)
out_file = open(
os.path.join(args.output_dir,
sub_dir_path_1 + "_Det_ffasta.txt"), "wb")
for img_idx, sub_file_path_2 in enumerate(
sub_file_path_list_2):
im = cv2.imread(
os.path.join(root_dir_path_2, sub_file_path_2))
timers = defaultdict(Timer)
t = time.time()
if (img_idx + 1) % 1000 == 0:
sys.stdout.write(
"\rFinish {} images\n".format(img_idx + 1))
sys.stdout.flush()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
if isinstance(cls_boxes, list):
cls_boxes, cls_segms, cls_keyps, classes = vis_utils.convert_from_cls_format(
cls_boxes, cls_segms, cls_keyps)
if cls_boxes is None or cls_boxes.shape[0] == 0:
continue
obj_idx = 0
for cls_box, cls in zip(cls_boxes, classes):
if int(cls) != 3 and int(cls) != 6:
continue
out_file.write("{},{},{},{},{},{},{}\n".format(
img_idx + 1, obj_idx + 1, cls_box[0],
cls_box[1], cls_box[2] - cls_box[0],
cls_box[3] - cls_box[1], cls_box[4]))
obj_idx += 1
out_file.close()
print("Finish {} / {} of video sequences".format(
i + 1, len(sub_dir_path_list_1)))
break
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
out_name = os.path.join(
args.output_dir, '{}'.format(os.path.basename(im_name) + '.pdf')
)
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers
)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)'
)
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2
)
def predict(cfg_path, weights_path):
logger = logging.getLogger(__name__)
merge_cfg_from_file(cfg_path)
cfg.NUM_GPUS = 1
gpu_id = 0 #指定GPU
weights_path = cache_url(weights_path, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(weights_path, gpu_id)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
#dummy_coco_dataset = dummy_datasets.get_illbuild_class11_dataset()
print("model is ok")
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
out_name = os.path.join(
args.output_dir, '{}'.format(os.path.basename(im_name) + '.jpg'))
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(gpu_id):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)')
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2)
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = (dummy_datasets.get_vg3k_dataset() if args.use_vg3k
else dummy_datasets.get_coco_dataset())
if args.im_or_folder == 'train0':
#im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
im_list = cPickle.load(open('train_list.pkl'))
pre_folder = '/mnt/storage/jialinwu/detectron/detectron/datasets/data/coco/coco_train2014'
im_list = im_list[:40000]
elif args.im_or_folder == 'train1':
im_list = cPickle.load(open('train_list.pkl'))
pre_folder = '/mnt/storage/jialinwu/detectron/detectron/datasets/data/coco/coco_train2014'
im_list = im_list[40000:]
elif args.im_or_folder == 'val':
im_list = cPickle.load(open('val_list.pkl'))
pre_folder = '/mnt/storage/jialinwu/detectron/detectron/datasets/data/coco/coco_val2014'
elif args.im_or_folder == 'test0':
im_list = cPickle.load(open('test_list.pkl'))
im_list = im_list[:40000]
pre_folder = '../ARNet/image_captioning/data/images/test2015'
elif args.im_or_folder == 'test1':
im_list = cPickle.load(open('test_list.pkl'))
im_list = im_list[40000:]
pre_folder = '../ARNet/image_captioning/data/images/test2015'
for i in tqdm(range(len(im_list))):
im_name = pre_folder + '/' + im_list[i]
if im_name[-4:] != '.jpg':
continue
im = cv2.imread(im_name)
timers = defaultdict(Timer)
imgid = int(im_list[i][:-4].split('_')[-1])
if args.im_or_folder == 'val':
save_name = '/mnt/storage/jialinwu/seg_every_thing/npz_features/coco_val2014/%d.npz' % imgid
else:
save_name = '/mnt/storage/jialinwu/seg_every_thing/npz_features/coco_train2014/%d.npz' % imgid
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, save_name, timers=timers)
'''
def main(args):
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
if args.info is None and args.scene_dir is None and args.video_file is not None: # video only
_main(args, model)
elif args.info is None and args.scene_dir is not None and args.video_file is None: # scene dir only
derived_dir = os.path.join(args.scene_dir, 'derived')
for cam_dir in glob.glob(os.path.join(derived_dir, 'cam_*')):
video_file = glob.glob(os.path.join(cam_dir, '*.mp4'))[0]
args.video_file = video_file
_main(args, model)
elif args.info is not None and args.scene_dir is None and args.video_file is None: # info only
_main(args, model)
else:
raise ValueError(
'Must specify either scene-dir or video_file not both.')
box = boxs_list[i]
drawBoxOnImg(im, box[1], box[2], box[3], box[4], 0, 0, frame)
frame += 1
workspace.GlobalInit(['caffe2', '--caffe2_log_level=0'])
utils.logging.setup_logging(__name__)
logger = logging.getLogger(__name__)
#args = parse_args()
# args.cfg = '/detectron/configs/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_2x.yaml'
# args.weights = '/data/workspace/fbdet/models/mask_rcnn_R_101_FPN_2x/model_final.pkl'
cfg_path = '/detectron/configs/12_2017_baselines/e2e_faster_rcnn_X-101-64x4d-FPN_1x.yaml'
weights_path = '/data/workspace/fbdet/models/X_101_64x4d_FPN_faster/model_final.pkl'
merge_cfg_from_file(cfg_path)
cfg.NUM_GPUS = 1
weights1 = cache_url(weights_path, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(weights1)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
def pipeline_mask(im):
if im is None:
print("im is null!")
timers = defaultdict(Timer)
#t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
def main(args):
#just show box and mask
cfg_file = r'/home/twang/Documents/detectron/configs/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_2x.yaml'
weights_file = r'/home/twang/Documents/detectron/model-weights/mask_rcnn_R-101-FPN_2x_model_final.pkl'
#==============================================================================
# #show keypoint
# cfg_file = r'/home/twang/Documents/detectron/model-weights/e2e_KeyPoint_Mask_RCNN/e2e_keypoint_rcnn_X-101-64x4d-FPN_1x.yaml'
# weights_file = r'/home/twang/Documents/detectron/model-weights/e2e_KeyPoint_Mask_RCNN/X-101-64x4d-FPN_1x.pkl'
#
#==============================================================================
merge_cfg_from_file(cfg_file)
cfg.NUM_GPUS = 2
weights = cache_url(weights_file, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
Person_Gender_CKPT = '/home/twang/Documents/tensorflow-classfication2/train-slim/trained-models/airport_gender3/training_all_layers/vgg_16_750000'
Person_Gender_LABELS = os.path.join(
'/home/twang/Documents/tensorflow-classfication2/gender-airport-entrance',
'labels.txt')
Person_Gender_category_index = load_labels(Person_Gender_LABELS)
Person_Gender_gpu_option = tf.GPUOptions(
per_process_gpu_memory_fraction=0.2)
Gender = Person_Gender(Person_Gender_CKPT, Person_Gender_category_index,
Person_Gender_gpu_option)
video_dir = args.video_dir
cap = cv2.VideoCapture(video_dir)
video_box_mask = cv2.VideoWriter(
'out_box_mask.mp4', cv2.VideoWriter_fourcc('M', 'J', 'P', 'G'), 5,
(1280, 720))
while cap.isOpened():
t1 = time.time()
ret, frame = cap.read()
if not ret:
break
frame = cv2.resize(frame, dsize=(1280, 720))
timers = defaultdict(Timer)
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, timers=timers)
thresh = 0.7
show_box = True
show_class = True
crop_person = True
filter_big_box = True
dataset = dummy_coco_dataset
frame_for_person_crop = frame.copy()
frame_for_box = frame.copy()
frame_for_mask = frame.copy()
frame_for_both = frame.copy()
"""Constructs a numpy array with the detections visualized."""
if isinstance(cls_boxes, list):
boxes, segms, keypoints, classes = convert_from_cls_format(
cls_boxes, cls_segms, cls_keyps)
if boxes is None or boxes.shape[0] == 0 or max(boxes[:, 4]) < thresh:
return frame
if segms is not None and len(segms) > 0:
masks = mask_util.decode(segms)
color_list = colormap()
color_list_selected = color_list[0:5]
else:
color_list = colormap()
color_list_selected = color_list[0:5]
# Display in largest to smallest order to reduce occlusion
areas = (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
sorted_inds = np.argsort(-areas)
for i in sorted_inds:
bbox = boxes[i, :4]
score = boxes[i, -1]
if score < thresh:
continue
# show class (person, backpack, handbag, suitcase)
class_default = ['person', 'backpack', 'handbag', 'suitcase']
if show_class:
class_str, class_text = get_class_string(
classes[i], score, dataset)
if class_text in class_default:
frame = vis_class(frame, (bbox[0], bbox[1] - 2), class_str)
frame_for_both = vis_class(frame_for_both,
(bbox[0], bbox[1] - 2),
class_str)
#filter big box
if filter_big_box:
aspect_ratio = (bbox[2] - bbox[0]) / (bbox[3] -
bbox[1])
if aspect_ratio < 1.5:
#show bounding box
if show_box:
frame = vis_bbox(frame,
(bbox[0], bbox[1], bbox[2] -
bbox[0], bbox[3] - bbox[1]))
frame_for_box = vis_bbox(
frame_for_box,
(bbox[0], bbox[1], bbox[2] - bbox[0],
bbox[3] - bbox[1]))
frame_for_both = vis_bbox(
frame_for_both,
(bbox[0], bbox[1], bbox[2] - bbox[0],
bbox[3] - bbox[1]))
# crop each person box, recognize gender
if crop_person and class_text == 'person':
(x1, y1, w, h) = (int(bbox[0]), int(bbox[1]),
int(bbox[2] - bbox[0]),
int(bbox[3] - bbox[1]))
x2 = x1 + w
y2 = y1 + h
cropped = frame_for_person_crop[y1:y2, x1:x2]
cv2.imwrite(
'/home/twang/Documents/detectron/temp/Hk-demo/genderwrite.jpg',
cropped)
gender_frame_saved = '/home/twang/Documents/detectron/temp/Hk-demo/genderwrite.jpg'
prob = Gender.predict(gender_frame_saved)
if prob[0] > prob[1]:
#show class name and probality
frame_for_both = vis_class(
frame_for_both, (x1, y1 + 10),
"Female:%.1f" % (prob[0]))
#show female mask
color_mask_female = color_list_selected[
0, 0:3]
frame_for_both = vis_mask(
frame_for_both, masks[..., i],
color_mask_female)
frame_for_box = vis_class(
frame_for_box, (x1, y1 + 10),
"Female:%.1f" % (prob[0]))
frame_for_box = vis_mask(
frame_for_box, masks[..., i],
color_mask_female)
else:
#show class name and probality
frame_for_both = vis_class(
frame_for_both, (x1, y1 + 10),
"Male:%.1f" % (prob[1]))
#show male mask
color_mask_male = color_list_selected[1,
0:3]
frame_for_both = vis_mask(
frame_for_both, masks[..., i],
color_mask_male)
frame_for_box = vis_class(
frame_for_box, (x1, y1 + 10),
"Male:%.1f" % (prob[1]))
frame_for_box = vis_mask(
frame_for_box, masks[..., i],
color_mask_male)
#show mask, different other category has deifferent color
if segms is not None and len(segms) > i:
if class_text == 'backpack':
color_mask = color_list_selected[2, 0:3]
frame_for_both = vis_mask(
frame_for_both, masks[..., i],
color_mask)
elif class_text == 'handbag':
color_mask = color_list_selected[3, 0:3]
frame_for_both = vis_mask(
frame_for_both, masks[..., i],
color_mask)
elif class_text == 'suitcase':
color_mask = color_list_selected[4, 0:3]
frame_for_both = vis_mask(
frame_for_both, masks[..., i],
color_mask)
else:
#show bounding box
if show_box:
frame = vis_bbox(frame,
(bbox[0], bbox[1], bbox[2] -
bbox[0], bbox[3] - bbox[1]))
# crop each box
if crop_person and class_text == 'person':
(x1, y1, w, h) = (int(bbox[0]), int(bbox[1]),
int(bbox[2] - bbox[0]),
int(bbox[3] - bbox[1]))
x2 = x1 + w
y2 = y1 + h
cropped = frame_for_person_crop[y1:y2, x1:x2]
cv2.imwrite(
'/home/twang/Documents/detectron/temp/Hk-demo/genderwrite.jpg',
cropped)
gender_frame_saved = '/home/twang/Documents/detectron/temp/Hk-demo/genderwrite.jpg'
prob = Gender.predict(gender_frame_saved)
if prob[0] > prob[1]:
cv2.putText(frame,
"Female:%.1f" % (prob[0] * 100),
(x1, y1 + 10),
cv2.FONT_HERSHEY_SIMPLEX,
0.35,
_GRAY,
lineType=cv2.LINE_AA)
else:
cv2.putText(frame,
"Male:%.1f" % (prob[1] * 100),
(x1, y1 + 10),
cv2.FONT_HERSHEY_SIMPLEX,
0.35,
_GRAY,
lineType=cv2.LINE_AA)
t2 = time.time()
durr = float(t2 - t1)
fps = 1.0 / durr
#cv2.putText(frame,"fps:%.3f"%fps,(20,20),4, 0.5, (0, 255, 0), 1, cv2.LINE_AA)
cv2.imshow('Detection using and box mask', frame_for_both)
video_box_mask.write(frame_for_both)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
video_box_mask.release()
cv2.destroyAllWindows()
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
start = timeit.default_timer()
# extract bboxes from bottom-up attention model
image_bboxes = {}
count = 0
if not os.path.exists(args.output_dir):
os.makedirs(args.output_dir)
results = {}
if os.path.isdir(args.im_or_folder):
# (legacy) the order of list_of_ids has to be consistent with dic_anet.json generated
# from the GVD prepro script for correct data loading
# target_ids = set(os.listdir(args.im_or_folder)) # temp, when a target split is given
# list_of_folder = [i['id'] for i in json.load(open(args.list_of_ids))['videos'] \
# if i['id'] in target_ids]
list_of_folder = [
i['id'] for i in json.load(open(args.list_of_ids))['videos']
]
else:
list_of_folder = []
N = len(list_of_folder)
print('Number of segments to generate proposals for: ', N)
fpv = 10
dets_labels = np.zeros((N, fpv, 100, 7))
dets_num = np.zeros((N, fpv))
nms_num = np.zeros((N, fpv))
hw = np.zeros((N, 2))
for i, folder_name in enumerate(list_of_folder):
dets_feat = []
for j in range(fpv):
im_name = os.path.join(args.im_or_folder, folder_name,
str(j + 1).zfill(2) + args.image_ext)
im = cv2.imread(im_name)
try:
result = get_detections_from_im(cfg, model, im, '', '',
args.feat_name,
args.min_bboxes,
args.max_bboxes)
except:
print('missing frame: ', im_name)
num_frm = j
break
height, width, _ = im.shape
hw[i, 0] = height
hw[i, 1] = width
# store results
num_proposal = result['boxes'].shape[0]
proposals = np.concatenate(
(result['boxes'], np.ones((num_proposal, 1)) * j,
np.expand_dims(result['object'], axis=1),
np.expand_dims(result['obj_prob'], axis=1)),
axis=1)
dets_feat.append(result['region_feat'].squeeze())
dets_labels[i, j, :num_proposal] = proposals
dets_num[i, j] = num_proposal
nms_num[i, j] = num_proposal # for now, treat them the same
# save features to individual npy files
feat_output_file = os.path.join(args.output_dir, folder_name + '.npy')
if len(dets_feat) > 0:
dets_feat = np.stack(dets_feat)
print('Processed clip {}, feature shape {}'.format(
folder_name, dets_feat.shape))
np.save(feat_output_file, dets_feat)
else:
print('Empty feature file! Skipping {}...'.format(folder_name))
count += 1
if count % 10 == 0:
end = timeit.default_timer()
epoch_time = end - start
print('process {:d} videos after {:.1f} s'.format(
count, epoch_time))
f = h5py.File(args.det_output_file, "w")
f.create_dataset("dets_labels", data=dets_labels.view(N, -1, 7))
f.create_dataset("dets_num", data=dets_num.sum(axis=-1))
f.create_dataset("nms_num", data=nms_num.sum(axis=-1))
f.create_dataset("hw", data=hw)
f.close()
def main():
cfg_file = r'/home/twang/Documents/detectron/configs/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_2x.yaml'
weights_file = r'/home/twang/Documents/detectron/model-weights/mask_rcnn_R-101-FPN_2x_model_final.pkl'
merge_cfg_from_file(cfg_file)
cfg.NUM_GPUS = 1
weights = cache_url(weights_file, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
video_dir = '/media/network_shared_disk/WangTao/test_video/KLA_airport/Entrance_Peak_Hour.avi'
cap = cv2.VideoCapture(video_dir)
print(*'MJPG')
fourcc = cv2.VideoWriter_fourcc(*'MJPG')
video_output = cv2.VideoWriter("out.mp4", fourcc, 5, (1280, 720))
while cap.isOpened():
t1 = time.time()
ret, frame = cap.read()
if not ret:
break
frame = cv2.resize(frame, dsize=(1280, 720))
timers = defaultdict(Timer)
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, timers=timers)
thresh = 0.7
show_box = True
show_box = True
show_class = True
crop_person = True
dataset = dummy_coco_dataset
frame_for_person_crop = frame.copy()
frame_for_mask = frame.copy()
"""Constructs a numpy array with the detections visualized."""
if isinstance(cls_boxes, list):
boxes, segms, keypoints, classes = convert_from_cls_format(
cls_boxes, cls_segms, cls_keyps)
if boxes is None or boxes.shape[0] == 0 or max(boxes[:, 4]) < thresh:
return frame
if segms is not None and len(segms) > 0:
masks = mask_util.decode(segms)
color_list = colormap()
mask_color_id = 0
# Display in largest to smallest order to reduce occlusion
areas = (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
sorted_inds = np.argsort(-areas)
for i in sorted_inds:
bbox = boxes[i, :4]
score = boxes[i, -1]
if score < thresh:
continue
# show class (person, backpack, handbag, suitcase)
class_default = ['person', 'backpack', 'handbag', 'suitcase']
if show_class:
class_str, class_text = get_class_string(
classes[i], score, dataset)
if class_text in class_default:
frame = vis_class(frame, (bbox[0], bbox[1] - 2), class_str)
#show bounding box
if show_box:
frame = vis_bbox(frame,
(bbox[0], bbox[1], bbox[2] - bbox[0],
bbox[3] - bbox[1]))
# show mask
if segms is not None and len(segms) > i:
color_mask = color_list[mask_color_id %
len(color_list), 0:3]
mask_color_id += 1
frame_for_mask = vis_mask(frame_for_mask,
masks[..., i], color_mask)
t2 = time.time()
durr = float(t2 - t1)
fps = 1.0 / durr
#cv2.putText(frame,"fps:%.3f"%fps,(20,20),4, 0.5, (0, 255, 0), 1, cv2.LINE_AA)
cv2.imshow('Detection using box', frame)
cv2.imshow('Detection using mask', frame_for_mask)
video_output.write(frame)
#video_mask.write(frame_for_mask)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
video_output.release()
cv2.destroyAllWindows()
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.video_name):
print("video_name", args.video_name)
else:
print("video is not existing")
cap = cv2.VideoCapture(args.video_name)
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break
frame = cv2.resize(frame, dsize=(1280, 720))
timers = defaultdict(Timer)
t1 = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t1))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
frame = vis_utils.vis_one_image_opencv(frame,
cls_boxes,
segms=cls_segms,
keypoints=cls_keyps,
thresh=0.8,
kp_thresh=2,
show_box=True,
dataset=dummy_coco_dataset,
show_class=True)
t2 = time.time()
durr = float(t2 - t1)
fps = 1.0 / durr
cv2.putText(frame, "fps:%.3f" % fps, (20, 20), 4, 0.5, (0, 255, 0), 1,
cv2.LINE_AA)
cv2.imshow('Detection', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
cv2.destroyAllWindows()
def main(args):
datasetName = 'fashion_seg_val' #'furniture_val'
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 2
vis = True #False
shuffleList = False #True
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
if args.cls_thrsh_file is not None:
class_thresholds = {
l.split('\t')[0]: float(l.rstrip().split('\t')[1])
for l in open(args.cls_thrsh_file, 'r').readlines()
}
print(class_thresholds)
else:
class_thresholds = None
#dummy_coco_dataset = dummy_datasets.get_coco_dataset()
dataset = JsonDataset(datasetName)
if args.im_list is None:
im_list = glob.glob(args.im_or_folder + '/*.' + args.image_ext)
im_list = [osp.basename(n) for n in im_list]
else:
im_list = [
l.rstrip() + '.jpg' for l in open(args.im_list, 'r').readlines()
]
if shuffleList:
from random import shuffle
shuffle(im_list)
checkMkdir(args.output_dir)
#outTable = osp.join(args.output_dir, 'HF_CT_Measurement_Detected_Boxes.tsv')
#with open(outTable,'wb') as fout:
for i, im_name in enumerate(im_list):
output_name = os.path.basename(im_name) + '.png'
outFileName = os.path.join(args.output_dir, output_name)
if osp.exists(outFileName):
print("{} exists! continue".format(outFileName))
continue
imgFileName = osp.join(args.im_or_folder, im_name)
print(imgFileName)
im = cv2.imread(imgFileName)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
#for k, v in timers.items():
# logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if vis:
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dataset, #dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2)
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
os.makedirs(args.output_dir, exist_ok=True)
base_name = os.path.splitext(os.path.basename(im_name))[0]
out_image = os.path.join(args.output_dir,
'{}'.format(base_name + '.png'))
out_data = os.path.join(args.output_dir,
'{}'.format(base_name + '.pickle'))
if os.path.isfile(out_image) and os.path.isfile(out_data):
# logger.info('Already processed {}, skipping'.format(im_name))
continue
else:
logger.info('Processing {} -> {}'.format(im_name, out_image))
im = cv2.imread(im_name)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)')
if not os.path.isfile(out_data):
with open(out_data, 'wb') as f:
pickle.dump(
{
'boxes': cls_boxes,
'segmentations': cls_segms,
'keypoints': cls_keyps
}, f)
if not os.path.isfile(out_image):
logging.info('Visualizing %s', out_image)
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
base_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2,
dpi=300,
ext='png')
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
# get the weight path
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
# glob.iglob will generate a iterator that has same elements as that in glob.glob
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
out_name = os.path.join(
args.output_dir, '{}'.format(os.path.basename(im_name) + '.pdf'))
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)')
# save as pdf using matplotlib.pyplot
proc_im = vis_utils.vis_one_image_opencv(im,
cls_boxes,
dataset=dummy_coco_dataset,
thresh=0.7,
show_box=True,
show_class=True)
cv2.imshow('img', proc_im)
cv2.waitKey(0)
"""
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2
)
"""
cv2.destroyWindow('img')
def main(args):
cfg_file = r'/home/twang/Documents/detectron/configs/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_2x.yaml'
weights_file = r'/home/twang/Documents/detectron/model-weights/mask_rcnn_R-101-FPN_2x_model_final.pkl'
video_dir = args.video_dir
print("video_dir", video_dir)
video_name = os.path.basename(video_dir)
video_name = os.path.splitext(video_name)[0]
print("video_name", video_name)
directory_box = os.path.join(
os.path.join(r"/home/twang/Documents/HK-person", video_name), 'box')
print("directory_box", directory_box)
os.makedirs(directory_box)
directory_mask = os.path.join(
os.path.join(r"/home/twang/Documents/HK-person", video_name), 'mask')
print("directory_mask", directory_mask)
os.makedirs(directory_mask)
merge_cfg_from_file(cfg_file)
cfg.NUM_GPUS = 1
weights = cache_url(weights_file, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg()
model = infer_engine.initialize_model_from_cfg(weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
cap = cv2.VideoCapture(video_dir)
count = 0
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break
frame = cv2.resize(frame, dsize=(1280, 720))
total_frame = cap.get(cv2.CAP_PROP_FRAME_COUNT)
current_frame = int(cap.get(cv2.CAP_PROP_POS_FRAMES))
Frame_step = 5
if current_frame + Frame_step < total_frame:
cap.set(cv2.CAP_PROP_POS_FRAMES, current_frame + Frame_step)
timers = defaultdict(Timer)
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, frame, None, timers=timers)
thresh = 0.9
crop_box = True
dataset = dummy_coco_dataset
frame_for_box_crop = frame.copy()
frame_for_mask = frame.copy()
"""Constructs a numpy array with the detections visualized."""
if isinstance(cls_boxes, list):
boxes, segms, keypoints, classes = convert_from_cls_format(
cls_boxes, cls_segms, cls_keyps)
if boxes is None or boxes.shape[0] == 0 or max(boxes[:,
4]) < thresh:
return frame
if segms is not None and len(segms) > 0:
masks = mask_util.decode(segms)
color_list = colormap()
mask_color_id = 0
# Display in largest to smallest order to reduce occlusion
areas = (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
sorted_inds = np.argsort(-areas)
for i in sorted_inds:
bbox = boxes[i, :4]
score = boxes[i, -1]
if score < thresh:
continue
# crop each box
if crop_box:
#frame = vis_bbox(frame, (bbox[0], bbox[1], bbox[2] - bbox[0], bbox[3] - bbox[1]))
(x1, y1, w, h) = (int(bbox[0]), int(bbox[1]),
int(bbox[2] - bbox[0]),
int(bbox[3] - bbox[1]))
x2 = x1 + w
y2 = y1 + h
cropped = frame_for_box_crop[y1:y2, x1:x2]
cv2.imwrite(
"%s/person_Frame%i_%i.png" %
(directory_box, current_frame, i), cropped)
# crop each mask
if segms is not None and len(segms) > i:
color_mask = color_list[mask_color_id % len(color_list),
0:3]
mask_color_id += 1
#frame = vis_mask(frame, masks[..., i], color_mask)
(x1, y1, w, h) = (int(bbox[0]), int(bbox[1]),
int(bbox[2] - bbox[0]),
int(bbox[3] - bbox[1]))
x2 = x1 + w
y2 = y1 + h
cropped_mask = masks[..., i]
cropped_mask = cropped_mask[y1:y2, x1:x2]
cropped_img = frame_for_mask[y1:y2, x1:x2]
cropped_img = vis_mask(cropped_img, cropped_mask,
color_mask)
cv2.imwrite(
"%s/person_Mask_Frame%i_%i.png" %
(directory_mask, current_frame, i), cropped_img)
count += 1
print("done:%i" % count)
else:
pass
cap.release()
cv2.destroyAllWindows()
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
dummy_coco_dataset = dummy_datasets.get_coco_dataset()
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' + args.image_ext)
else:
im_list = [args.im_or_folder]
for i, im_name in enumerate(im_list):
out_name = os.path.join(
args.output_dir, '{}'.format(os.path.basename(im_name) + '.pdf'))
logger.info('Processing {} -> {}'.format(im_name, out_name))
im = cv2.imread(im_name)
if im is None:
continue
timers = defaultdict(Timer)
t = time.time()
with c2_utils.NamedCudaScope(0):
cls_boxes, cls_segms, cls_keyps = infer_engine.im_detect_all(
model, im, None, timers=timers)
logger.info('Inference time: {:.3f}s'.format(time.time() - t))
for k, v in timers.items():
logger.info(' | {}: {:.3f}s'.format(k, v.average_time))
if i == 0:
logger.info(
' \ Note: inference on the first image will be slower than the '
'rest (caches and auto-tuning need to warm up)')
lists = vis_utils.vis_one_image2(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2)
class_name = []
class_prob = []
if lists is not None:
for x in lists:
class_name.append(x.split(' ')[0])
class_prob.append(x.split(' ')[1])
if (len(class_name) == len(set(class_name))):
if (class_name[0] == 'apple'):
lists.insert(0, '1')
else:
lists.insert(0, '0')
else:
lists.insert(0, '0')
else:
lists = []
lists.insert(0, '2')
lists.insert(0, os.path.splitext(os.path.basename(im_name))[0])
with open(os.path.join(args.output_dir, 'output.csv'),
'a') as resultFile:
wr = csv.writer(resultFile, dialect='excel')
wr.writerow(lists)
vis_utils.vis_one_image(
im[:, :, ::-1], # BGR -> RGB for visualization
im_name,
args.output_dir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dummy_coco_dataset,
box_alpha=0.3,
show_class=True,
thresh=0.7,
kp_thresh=2)
def main(args):
logger = logging.getLogger(__name__)
merge_cfg_from_file(args.cfg)
cfg.NUM_GPUS = 1
args.weights = cache_url(args.weights, cfg.DOWNLOAD_CACHE)
assert_and_infer_cfg(cache_urls=False)
model = infer_engine.initialize_model_from_cfg(args.weights)
start = timeit.default_timer()
# extract region bboxes and features from pre-trained models
count = 0
if not os.path.exists(args.box_output_dir):
os.makedirs(args.box_output_dir)
if not os.path.exists(args.featcls_output_dir):
os.makedirs(args.featcls_output_dir)
results = {}
with h5py.File(os.path.join(args.box_output_dir, args.output_file_prefix+'_bbox'+args.proc_split+'.h5'), "w") as f, \
h5py.File(os.path.join(args.featcls_output_dir, args.output_file_prefix+'_feat'+args.proc_split+'.h5'), "w") as f_feat, \
h5py.File(os.path.join(args.featcls_output_dir, args.output_file_prefix+'_cls'+args.proc_split+'.h5'), "w") as f_cls:
if args.dataset in ('COCO', 'Flickr30k'):
if os.path.isdir(args.im_or_folder):
im_list = glob.iglob(args.im_or_folder + '/*.' +
args.image_ext)
elif args.dataset == 'CC':
valid_ids = json.load(
open('/mnt/dat/CC/annotations/cc_valid_jpgs.json')
) # some images are broken. hard coded for now
im_list = valid_ids.keys()
print('number of valid CC images {}'.format(len(im_list)))
elif args.dataset == 'SBU':
valid_ids = json.load(
open('/z/dat/VLP/dat/SBU/annotations/sbu_valid_jpgs.json')
) # some images are broken. hard coded for now
im_list = valid_ids.keys()
print('number of valid SBU images {}'.format(len(im_list)))
for i, im_name in enumerate(im_list):
im_base_name = os.path.basename(im_name)
image_id = im_base_name
if image_id[-4 - len(args.proc_split):-4] == args.proc_split:
im_name = os.path.join(args.im_or_folder, image_id)
print(im_name)
im = cv2.imread(im_name)
result = get_detections_from_im(cfg, model, im, image_id, '',
args.feat_name,
args.min_bboxes,
args.max_bboxes)
# store results
proposals = np.concatenate((result['boxes'], np.expand_dims(result['object'], axis=1) \
.astype(np.float32), np.expand_dims(result['obj_prob'], axis=1)), axis=1)
f.create_dataset(image_id[:-4],
data=proposals.astype(args.data_type))
f_feat.create_dataset(
image_id[:-4],
data=result['region_feat'].squeeze().astype(
args.data_type))
f_cls.create_dataset(image_id[:-4],
data=result['cls_prob'].astype(
args.data_type))
count += 1
if count % 10 == 0:
end = timeit.default_timer()
epoch_time = end - start
print('process {:d} images after {:.1f} s'.format(
count, epoch_time))