VideoReader

class torchvision.io.VideoReader(src: str, stream: str = 'video', num_threads: int = 0)

Fine-grained video-reading API. Supports frame-by-frame reading of various streams from a single video container. Much like previous video_reader API it supports the following backends: video_reader, pyav, and cuda. Backends can be set via torchvision.set_video_backend function.

Warning

The VideoReader class is in Beta stage, and backward compatibility is not guaranteed.

Example

The following examples creates a VideoReader object, seeks into 2s point, and returns a single frame:

import torchvision
video_path = "path_to_a_test_video"
reader = torchvision.io.VideoReader(video_path, "video")
reader.seek(2.0)
frame = next(reader)

VideoReader implements the iterable API, which makes it suitable to using it in conjunction with itertools for more advanced reading. As such, we can use a VideoReader instance inside for loops:

reader.seek(2)
for frame in reader:
    frames.append(frame['data'])
# additionally, `seek` implements a fluent API, so we can do
for frame in reader.seek(2):
    frames.append(frame['data'])

With itertools, we can read all frames between 2 and 5 seconds with the following code:

for frame in itertools.takewhile(lambda x: x['pts'] <= 5, reader.seek(2)):
    frames.append(frame['data'])

and similarly, reading 10 frames after the 2s timestamp can be achieved as follows:

for frame in itertools.islice(reader.seek(2), 10):
    frames.append(frame['data'])

Note

Each stream descriptor consists of two parts: stream type (e.g. ‘video’) and a unique stream id (which are determined by the video encoding). In this way, if the video container contains multiple streams of the same type, users can access the one they want. If only stream type is passed, the decoder auto-detects first stream of that type.

Parameters:

• src (string, bytes object, or tensor) – The media source. If string-type, it must be a file path supported by FFMPEG. If bytes, should be an in-memory representation of a file supported by FFMPEG. If Tensor, it is interpreted internally as byte buffer. It must be one-dimensional, of type torch.uint8.

• stream (string, optional) – descriptor of the required stream, followed by the stream id, in the format {stream_type}:{stream_id}. Defaults to "video:0". Currently available options include ['video', 'audio']

• num_threads (int, optional) – number of threads used by the codec to decode video. Default value (0) enables multithreading with codec-dependent heuristic. The performance will depend on the version of FFMPEG codecs supported.

Examples using VideoReader:

Optical Flow: Predicting movement with the RAFT model

Optical Flow: Predicting movement with the RAFT model

Video API

Video API

get_metadata() → Dict[str, Any]

Returns video metadata

Returns:

dictionary containing duration and frame rate for every stream

Return type:

(dict)

seek(time_s: float, keyframes_only: bool = False) → VideoReader

Seek within current stream.

Parameters:

• time_s (float) – seek time in seconds

• keyframes_only (bool) – allow to seek only to keyframes

Note

Current implementation is the so-called precise seek. This means following seek, call to next() will return the frame with the exact timestamp if it exists or the first frame with timestamp larger than time_s.

set_current_stream(stream: str) → bool

Set current stream. Explicitly define the stream we are operating on.

Parameters:

stream (string) – descriptor of the required stream. Defaults to "video:0" Currently available stream types include ['video', 'audio']. Each descriptor consists of two parts: stream type (e.g. ‘video’) and a unique stream id (which are determined by video encoding). In this way, if the video container contains multiple streams of the same type, users can access the one they want. If only stream type is passed, the decoder auto-detects first stream of that type and returns it.

Returns:

True on success, False otherwise

Return type:

(bool)