MultimodalInput<T>

Models & Types · AiDotNet.Interfaces

Represents an input item for unified multimodal models.

Constructors

Constructor Summary
MultimodalInput()

Properties

Property Summary
Dictionary<string, object> Metadata { get; set; } Optional metadata about the input.
ModalityType Modality { get; set; } The modality type of this input.
int SequenceIndex { get; set; } Temporal ordering for sequential inputs.
string TextContent { get; set; } Optional text content (for text modality).

Methods

Method Summary
MultimodalInput<T> FromAudio(Vector<T> samples, int sampleRate, int sequenceIndex) Creates an audio input from waveform samples.
MultimodalInput<T> FromImage(Vector<T> pixels, int channels, int height, int width, int sequenceIndex) Creates an image input from pixel data.
MultimodalInput<T> FromText(string text, int sequenceIndex) Creates a text input for the multimodal model.
MultimodalInput<T> FromVideo(Vector<T> frames, int numFrames, int channels, int height, int width, double frameRate, int sequenceIndex) Creates a video input from frame data.