Last active
September 7, 2017 16:52
-
-
Save xkumiyu/9baf81ac40ee43f87bce80b991c969d2 to your computer and use it in GitHub Desktop.
Chainer Dataset of Videos
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import os | |
import numpy as np | |
import six | |
import cv2 | |
import chainer | |
def _read_video_as_array(path, dtype): | |
video = [] | |
cap = cv2.VideoCapture(path) | |
while(cap.isOpened()): | |
ret, frame = cap.read() | |
if ret: | |
video.append(frame) | |
else: | |
break | |
cap.release() | |
return np.asarray(video, dtype=dtype) | |
class VideoDataset(chainer.dataset.DatasetMixin): | |
def __init__(self, paths, root='.', dtype=np.float32): | |
if isinstance(paths, six.string_types): | |
with open(paths) as paths_file: | |
paths = [path.strip() for path in paths_file] | |
self._paths = paths | |
self._root = root | |
self._dtype = dtype | |
def __len__(self): | |
return len(self._paths) | |
def get_example(self, i): | |
""" | |
return video shape: (ch, frame, width, height) | |
""" | |
path = os.path.join(self._root, self._paths[i]) | |
video = _read_video_as_array(path, self._dtype) | |
return video.transpose(3, 0, 2, 1) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment