lib.video Module

Utilities for working with videos

lib.video.VIDEO_EXTENSIONS = ['.avi', '.flv', '.mkv', '.mov', '.mp4', '.mpeg', '.mpg', '.webm', '.wmv', '.ts', '.vob']

List of lowercase valid Video extensions with preceding period

class lib.video.VideoInfo(video_file: str, fast_count: bool = True, stream_index: int = 0, pts: list[int] | None = None, keyframes: list[int] | None = None)

Collects and stores information about video files

Parameters:

• video_file (str) – Full path to a video file

• fast_count (bool) – Whether to obtain the count of frames quickly, but inaccurately or slowly but accurately. If pts and keyframes are provided then the count will be derived from the provided pts file. Default: True

• stream_index (int) – The stream index to select from the video file. Default: 0

• pts (list[int] | None) – The Presentation Timestamps if available or None to retrieve from the video. Default: None

• keyframes (list[int] | None) – The keyframe frame indices if available or None to retrieve from the video. Default: None

property count: int

The number of frames in the video

property duration: int

The duration of the video file in seconds

property keyframes: npt.NDArray[np.int64]

The frame index of each key frame in the video

property keyframes_count: int

The number of keyframes that exist in the video

property pts: npt.NDArray[np.int64]

The Presentation Time Stamp for each frame in the video

class lib.video.VideoMux(source_video: str, destination_video: str, codec: Literal['libx264', 'libx265'], codec_parameters: dict[str, str], mux_audio: bool = True)

A basic muxer for muxing converted faceswap frames to a video file using the original video as a reference

Parameters:

• source_video (str) – The path to the source video to use as a reference for Audio and FPS

• destination_video (str) – The full path to save the final video to

• codec (T.Literal['libx264', 'libx265']) – The codec to use to encode the video

• codec_parameters (dict[str, str]) – The options to use for the codec

• mux_audio (bool) – True to mux order from the source video to the output

encode(image: npt.NDArray[np.uint8] | None) → None

Encode a frame to the video

Parameters:

image (npt.NDArray[np.uint8] | None) – The 3 channel BGR UINT8 image to encode to the video or None to finalize the video

Return type:

None

class lib.video.VideoReader(video_file: str, fast_count: bool = True, stream_index: int = 0, pts: list[int] | None = None, keyframes: list[int] | None = None)

A wrapper around pyAV that allows obtaining frames by frame index and iterating video files

Parameters:

• video_file (str) – Full path to a video file

• fast_count (bool) – Whether to obtain the count of frames quickly, but inaccurately or slowly but accurately. If pts and keyframes are provided then the count will be derived from the provided pts file. Default: True

• stream_index (int) – The stream index to select from the video file. Default: 0

• pts (list[int] | None) – The Presentation Timestamps if available or None to retrieve from the video. Default: None

• keyframes (list[int] | None) – The keyframe frame indices if available or None to retrieve from the video. Default: None

__next__() → VideoFrame

Obtain the next video frame object

Return type:

The next available video frame object

close() → None

Shut down the AV Container object

Return type:

None

get(index: int) → VideoFrame

Obtain the video frame at the given frame index

Parameters:

index (int) – The index number of the frame to retrieve

Return type:

The pyAV frame object for the given index

property info: VideoInfo

The metadata information for the video file

lib.video.check_for_video(input_location: str) → bool

Check whether the given input is a video file or a folder

Parameters:

input_location (str) – Full path to an input file

Returns:

bool

Return type:

‘True’ if input is a video ‘False’ if it is a folder.

Raises:

FaceswapError – If the given location is a file and does not have a valid video extension.

lib.video.count_frames(filename, fast=False)

Count the number of frames in a video file

There is no guaranteed accurate way to get a count of video frames without iterating through a video and decoding every frame.

count_frames() can return an accurate count (albeit fairly slowly) or a possibly less accurate count, depending on the fast parameter. A progress bar is displayed.

Parameters:

• filename (str) – Full path to the video to return the frame count from.

• fast (bool, optional) – Whether to count the frames without decoding them. This is significantly faster but accuracy is not guaranteed. Default: False.

Returns:

The number of frames in the given video file.

Return type:

int

Example

>>> filename = "/path/to/video.mp4"
>>> frame_count = count_frames(filename)

lib.video.validate_video_file(file_path: str) → str

Validates that a given file exists and is a valid video format

Parameters:

file_path (str) – The full path to the video file to validate

Return type:

The full expanded video file path

Raises:

FaceswapError – If the given video file is not valid

Functions

ceil(x, /)	Return the ceiling of x as an Integral.
check_for_video(input_location)	Check whether the given input is a video file or a folder
convert_to_secs(*args)	Convert time in hours, minutes, and seconds to seconds.
count_frames(filename[, fast])	Count the number of frames in a video file
get_module_objects(module)	Return a list of all public objects within the given module
parse_class_init(locals_dict)	Parse a locals dict from a class and return in a format suitable for logging :param locals_dict: A locals() dictionary from a newly initialized class
validate_video_file(file_path)	Validates that a given file exists and is a valid video format

Classes

FaceswapError	Faceswap Error for handling specific errors with useful information.
Fraction([numerator, denominator])	This class implements rational numbers.
VideoInfo(video_file[, fast_count, ...])	Collects and stores information about video files
VideoMux(source_video, destination_video, ...)	A basic muxer for muxing converted faceswap frames to a video file using the original video as a reference
VideoReader(video_file[, fast_count, ...])	A wrapper around pyAV that allows obtaining frames by frame index and iterating video files
deque	A list-like sequence optimized for data accesses near its endpoints.
tqdm(*_, **__)	Decorate an iterable object, returning an iterator which acts exactly like the original iterable, but prints a dynamically updating progressbar every time a value is requested.

Variables

VIDEO_EXTENSIONS	List of lowercase valid Video extensions with preceding period
annotations
logger	A standard logging.logger with additional "verbose" and "trace" levels added.

Class Inheritance Diagram