Open libingsy opened 5 years ago
I want to ask that i create two identical dataloaders, but the dataset gets different output and the same label, it make me confused. import numpy as np from coviar import load
video = '/data/libing/Data/hmdb51/mpeg4_videos/sword/_Kill_Bill__Uma_Thruman_sword_fight_with_Lucy_liu_sword_f_nm_np2_ri_med_0.mp4' num_frames = 126 gop_index = 0 gop_pos = 0
def get_frames(): len = 5 frames = np.zeros((len + 1, 256, 340, 3), dtype=np.uint8) for i in range(len): frames[i] = load(video, gop_index, gop_pos, 0, True)
return frames
def do_something(): frames = get_frames() frames2 = get_frames() c = frames - frames2 print(np.mean(c, axis=(1, 2, 3)))
I want to ask that i create two identical dataloaders, but the dataset gets different output and the same label, it make me confused. import numpy as np from coviar import load
video = '/data/libing/Data/hmdb51/mpeg4_videos/sword/_Kill_Bill__Uma_Thruman_sword_fight_with_Lucy_liu_sword_f_nm_np2_ri_med_0.mp4' num_frames = 126 gop_index = 0 gop_pos = 0
def get_frames(): len = 5 frames = np.zeros((len + 1, 256, 340, 3), dtype=np.uint8) for i in range(len): frames[i] = load(video, gop_index, gop_pos, 0, True)
print(frames[len, 0, 0, 0])
def do_something(): frames = get_frames() frames2 = get_frames() c = frames - frames2 print(np.mean(c, axis=(1, 2, 3)))