Masked Autoencoders that Listen

This paper studies a simple extension of image-based Masked Autoencoders (MAE) to self-supervised representation learning from audio spectrograms. Following the Transformer encoder-decoder design in MAE, our Audio-MAE first encodes audio spectrogram patches with a high masking ratio, feeding only the non-masked tokens through encoder layers. The decoder then re-orders and decodes the encoded context padded with mask tokens, in order to reconstruct the input spectrogram. We find it beneficial to incorporate local window attention in the decoder, as audio spectrograms are highly correlated in local time and frequency bands. We then fine-tune the encoder with a lower masking ratio on target datasets. Empirically, Audio-MAE sets new state-of-the-art performance on six audio and speech classification tasks, outperforming other recent models that use external supervised pre-training. The code and models will be at https://github.com/facebookresearch/AudioMAE.

PDF Abstract

Results from the Paper


 Ranked #1 on Speaker Identification on VoxCeleb1 (Accuracy metric)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Audio Classification AudioSet Audio-MAE (local, AS-2M) Test mAP 0.473 # 6
Audio Classification AudioSet Audio-MAE (global, AS-2M) Test mAP 0.468 # 10
Speaker Identification VoxCeleb1 AudioMAE (local) Accuracy 94.8 # 1
Speaker Identification VoxCeleb1 AudioMAE (global) Accuracy 94.1 # 3

Methods