AST: Audio Spectrogram Transformer

Code

• github.com/YuanGongND/ast
  OfficialIn paperpytorch★ 1,440
• github.com/nttcslab/composing-general-audio-repr
  pytorch★ 26
• github.com/pxaris/ccml
  pytorch★ 13
• github.com/cgaroufis/msspt
  tf★ 12
• github.com/pwc-1/Paper-8/tree/main/audio_spectrogram_transformer
  mindspore★ 0

Abstract

In the past decade, convolutional neural networks (CNNs) have been widely adopted as the main building block for end-to-end audio classification models, which aim to learn a direct mapping from audio spectrograms to corresponding labels. To better capture long-range global context, a recent trend is to add a self-attention mechanism on top of the CNN, forming a CNN-attention hybrid model. However, it is unclear whether the reliance on a CNN is necessary, and if neural networks purely based on attention are sufficient to obtain good performance in audio classification. In this paper, we answer the question by introducing the Audio Spectrogram Transformer (AST), the first convolution-free, purely attention-based model for audio classification. We evaluate AST on various audio classification benchmarks, where it achieves new state-of-the-art results of 0.485 mAP on AudioSet, 95.6% accuracy on ESC-50, and 98.1% accuracy on Speech Commands V2.

Tasks

Benchmark Results

Reproductions