VideoNeuralNetworkBase<T>

Base Classes ยท AiDotNet.Video

Base class for video-focused neural networks that can operate in both ONNX inference and native training modes.

Inheritance: NeuralNetworkBase<T> โ†’ VideoNeuralNetworkBase<T>

Known derived types: FrameInterpolationBase<T>, Mcnet<T>, OpticalFlowBase<T>, VideoDenoisingBase<T>, VideoInpaintingBase<T>, VideoStabilizationBase<T>, VideoSuperResolutionBase<T>

For Beginners

Video neural networks process sequences of image frames to perform tasks like super-resolution, frame interpolation, optical flow estimation, denoising, stabilization, and inpainting. This base class provides:

  • Support for pre-trained ONNX models (fast inference with existing models)
  • Full training capability from scratch (like other neural networks)
  • Frame preprocessing utilities (normalization, patch extraction)
  • Temporal context handling (multi-frame processing)

You can use derived classes in two ways:

  1. Load a pre-trained ONNX model for quick inference
  2. Build and train a new model from scratch

How It Works

This class extends NeuralNetworkBase to provide video-specific functionality while maintaining full integration with the AiDotNet neural network infrastructure.

Properties

Property Summary
ILossFunction<T> DefaultLossFunction { get; } Gets the default loss function for this model.
bool IsOnnxMode { get; } Gets whether this model is running in ONNX inference mode.
bool SupportsTraining { get; } Gets whether this network supports training.
int FrameHeight { get; set; } Gets or sets the expected frame height for this model.
int FrameWidth { get; set; } Gets or sets the expected frame width for this model.
int NumChannels { get; set; } Gets or sets the number of color channels expected by this model.
int NumFrames { get; set; } Gets or sets the number of frames this model processes at once.