Table of Contents

Class VideoAugmenterBase<T>

Namespace
AiDotNet.Augmentation.Video
Assembly
AiDotNet.dll

Base class for video data augmentations.

public abstract class VideoAugmenterBase<T> : AugmentationBase<T, ImageTensor<T>[]>, IAugmentation<T, ImageTensor<T>[]>

Type Parameters

T

The numeric type for calculations.

Inheritance
VideoAugmenterBase<T>
Implements
Derived
Inherited Members

Remarks

For Beginners: Video augmentation transforms sequences of frames to improve model robustness to temporal variations. It combines:

  • Temporal augmentations (time-based): cropping, reversing, speed changes
  • Spatial augmentations (frame-based): flips, rotations, color changes applied consistently across frames

Video data is represented as an array of ImageTensor frames.

Constructors

VideoAugmenterBase(double, double)

Initializes a new video augmentation.

protected VideoAugmenterBase(double probability = 1, double frameRate = 30)

Parameters

probability double

The probability of applying this augmentation (0.0 to 1.0).

frameRate double

The frame rate of the video in fps.

Properties

FrameRate

Gets or sets the frame rate of the video in frames per second.

public double FrameRate { get; set; }

Property Value

double

Remarks

Default: 30 fps (common for most video)

Other common values: 24 fps (film), 25 fps (PAL), 60 fps (high frame rate)

Methods

GetDuration(ImageTensor<T>[])

Gets the duration of the video in seconds.

protected double GetDuration(ImageTensor<T>[] frames)

Parameters

frames ImageTensor<T>[]

The video frames.

Returns

double

The duration in seconds.

GetFrameCount(ImageTensor<T>[])

Gets the number of frames in the video.

protected int GetFrameCount(ImageTensor<T>[] frames)

Parameters

frames ImageTensor<T>[]

The video frames.

Returns

int

The frame count.

GetParameters()

Gets the parameters of this augmentation.

public override IDictionary<string, object> GetParameters()

Returns

IDictionary<string, object>

A dictionary of parameter names to values.

ValidateFrameDimensions(ImageTensor<T>[])

Validates that all frames have the same dimensions.

protected void ValidateFrameDimensions(ImageTensor<T>[] frames)

Parameters

frames ImageTensor<T>[]

The video frames to validate.

Exceptions

ArgumentException

Thrown if frames have inconsistent dimensions.