def get_dataflow(coco_data_paths):
"""
This function initializes the tensorpack dataflow and serves generator
for training operation.
:param coco_data_paths: paths to the coco files: annotation file and folder with images
:return: dataflow object
"""
df = CocoDataFlow((368, 368), coco_data_paths)
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_sample)
df = PrefetchDataZMQ(df, nr_proc=4) #df = PrefetchData(df, 2, 1)
return df
def get_dataflow(annot_path, img_dir):
"""
This function initializes the tensorpack dataflow and serves generator
for training operation.
:param annot_path: path to the annotation file
:param img_dir: path to the images
:return: dataflow object
"""
df = CocoDataFlow((368, 368), annot_path, img_dir)
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_sample)
df = PrefetchDataZMQ(df, nr_proc=4) #df = PrefetchData(df, 2, 1)
return df
46,
meta.aug_joints,
1,
stride=8)
return [meta, mask_paf, mask_heatmap, pafmap, heatmap]
if __name__ == '__main__':
batch_size = 10
curr_dir = os.path.dirname(__file__)
annot_path = os.path.join(
curr_dir, '../dataset/annotations/pen_keypoints_validation.json')
img_dir = os.path.abspath(os.path.join(curr_dir, '../dataset/validation/'))
df = CocoDataFlow(
(368, 368), COCODataPaths(annot_path, img_dir)) #, select_ids=[1000])
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_debug_sample)
df = PrefetchData(df, nr_prefetch=2, nr_proc=1)
df.reset_state()
gen = df.get_data()
for g in gen:
show_image_mask_center_of_main_person(g)
#show_image_heatmap_paf(g)
df.reset_state()
return df
if __name__ == '__main__':
"""
Run this script to check speed of generating samples. Tweak the nr_proc
parameter of PrefetchDataZMQ. Ideally it should reflect the number of cores
in your hardware
"""
batch_size = 10
curr_dir = os.path.dirname(__file__)
annot_path = os.path.join(
curr_dir, '../dataset/annotations/person_keypoints_val2017.json')
img_dir = os.path.abspath(os.path.join(curr_dir, '../dataset/val2017/'))
df = CocoDataFlow(
(368, 368), COCODataPaths(annot_path, img_dir)) # , select_ids=[1000])
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_sample)
df = PrefetchDataZMQ(df, nr_proc=4)
df = BatchData(df, batch_size, use_list=False)
df = MapData(
df, lambda x: ([x[0], x[1], x[2]], [
x[3], x[4], x[3], x[4], x[3], x[4], x[3], x[4], x[3], x[4], x[3],
x[4]
]))
TestDataSpeed(df, size=100).start()
def get_dataflow(annot_path, img_dir, batch_size):
"""
This function initializes the tensorpack dataflow and serves generator
for training operation.
:param annot_path: path to the annotation file
:param img_dir: path to the images
:param batch_size: batch size
:return: dataflow object
"""
df = CocoDataFlow((368, 368), annot_path, img_dir)
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_sample)
df = PrefetchDataZMQ(df, nr_proc=4) #df = PrefetchData(df, 2, 1)
df = BatchData(df, batch_size, use_list=False)
df = MapData(
df, lambda x: ([x[0], x[1], x[2]], [
x[3], x[4], x[3], x[4], x[3], x[4], x[3], x[4], x[3], x[4], x[3],
x[4]
]))
df.reset_state()
return df
x[4],
x[3],
x[4],
]))
df.reset_state()
return df
if __name__ == '__main__':
"""
Run this script to check speed of generating samples. Tweak the nr_proc
parameter of PrefetchDataZMQ. Ideally it should reflect the number of cores
in your hardware
"""
batch_size = 10
curr_dir = os.path.dirname(__file__)
annot_path = os.path.join(curr_dir, '../dataset/my_person_keypoints.json')
img_dir = os.path.abspath(os.path.join(curr_dir, '../dataset/train_data/'))
df = CocoDataFlow((320, 320), annot_path, img_dir) #, select_ids=[1000])
df.prepare()
df = MapData(df, read_img)
df = MapData(df, gen_mask)
df = MapData(df, augment)
df = MapData(df, apply_mask)
df = MapData(df, build_sample)
df = PrefetchData(df, nr_proc=4)
df = BatchData(df, batch_size, use_list=False)
df = MapData(df, lambda x: ([x[0], x[1], x[2]], [x[3], x[4], x[3], x[4]]))
TestDataSpeed(df, size=100).start()