torchvision.io

The torchvision.io package provides functions for performing IO operations. They are currently specific to reading and writing video.

Video

torchvision.io.read_video(filename, start_pts=0, end_pts=None)

Reads a video from a file, returning both the video frames as well as the audio frames

Parameters

• filename (str) – path to the video file

• start_pts (int, optional) – the start presentation time of the video

• end_pts (int, optional) – the end presentation time

Returns

• vframes (Tensor[T, H, W, C]) – the T video frames

• aframes (Tensor[K, L]) – the audio frames, where K is the number of channels and L is the number of points

• info (Dict) – metadata for the video and audio. Can contain the fields video_fps (float) and audio_fps (int)

torchvision.io.read_video_timestamps(filename)

List the video frames timestamps.

Note that the function decodes the whole video frame-by-frame.

Parameters

filename (str) – path to the video file

Returns

• pts (List[int]) – presentation timestamps for each one of the frames in the video.

• video_fps (int) – the frame rate for the video

torchvision.io.write_video(filename, video_array, fps, video_codec='libx264', options=None)

Writes a 4d tensor in [T, H, W, C] format in a video file

Parameters

• filename (str) – path where the video will be saved

• video_array (Tensor[T, H, W, C]) – tensor containing the individual frames, as a uint8 tensor in [T, H, W, C] format

• fps (Number) – frames per second