AudioMAEOptions
Options & Configuration · AiDotNet.Audio.Classification
Configuration options for the Audio-MAE (Masked Autoencoders for Audio) model.
Inheritance: ModelOptions → AudioMAEOptions
For Beginners
Audio-MAE learns by playing “fill in the blanks” with spectrograms. It hides 80% of the audio picture and tries to reconstruct what was hidden. This forces the model to truly understand audio patterns. After this pre-training, it can be fine-tuned for classification with very little labeled data.
How It Works
Audio-MAE (Xu et al., NeurIPS 2022) applies the Masked Autoencoder framework to audio spectrograms for self-supervised pre-training. By masking 80% of spectrogram patches and reconstructing them, Audio-MAE learns rich audio representations that achieve 47.3% mAP on AudioSet and 97.0% on ESC-50 after fine-tuning.
References:
- Paper: “Masked Autoencoders that Listen” (Xu et al., NeurIPS 2022)
- Repository:
https://github.com/facebookresearch/AudioMAE
Constructors
| Constructor | Summary |
|---|---|
AudioMAEOptions() |
Properties
| Property | Summary |
|---|---|
OnnxModelOptions OnnxOptions { get; set; } |
Gets or sets ONNX runtime options. |
bool UseLocalAttention { get; set; } |
Gets or sets whether to use local attention in the encoder. |
double DropoutRate { get; set; } |
Gets or sets the dropout rate. |
double FeedForwardRatio { get; set; } |
Gets or sets the feed-forward dimension ratio. |
double LabelSmoothing { get; set; } |
Gets or sets the label smoothing factor. |
double LearningRate { get; set; } |
Gets or sets the initial learning rate. |
double MaskRatio { get; set; } |
Gets or sets the mask ratio for pre-training. |
double Threshold { get; set; } |
Gets or sets the confidence threshold. |
double WindowOverlap { get; set; } |
Gets or sets the window overlap ratio. |
double WindowSize { get; set; } |
Gets or sets the window size in seconds. |
int DecoderEmbeddingDim { get; set; } |
Gets or sets the decoder embedding dimension. |
int EncoderEmbeddingDim { get; set; } |
Gets or sets the encoder embedding dimension. |
int FMax { get; set; } |
Gets or sets the maximum frequency. |
int FMin { get; set; } |
Gets or sets the minimum frequency. |
int FftSize { get; set; } |
Gets or sets the FFT window size. |
int HopLength { get; set; } |
Gets or sets the hop length. |
int LocalAttentionWindow { get; set; } |
Gets or sets the local attention window size. |
int NumDecoderHeads { get; set; } |
Gets or sets the number of decoder attention heads. |
int NumDecoderLayers { get; set; } |
Gets or sets the number of decoder layers. |
int NumEncoderHeads { get; set; } |
Gets or sets the number of encoder attention heads. |
int NumEncoderLayers { get; set; } |
Gets or sets the number of encoder layers. |
int NumMels { get; set; } |
Gets or sets the number of mel bands. |
int PatchSize { get; set; } |
Gets or sets the patch size. |
int PatchStride { get; set; } |
Gets or sets the patch stride. |
int SampleRate { get; set; } |
Gets or sets the audio sample rate in Hz. |
int WarmUpEpochs { get; set; } |
Gets or sets the number of warm-up epochs. |
string ModelPath { get; set; } |
Gets or sets the path to a pre-trained ONNX model file. |
string[] CustomLabels { get; set; } |
Gets or sets custom event labels. |