LEAF: A Learnable Frontend for Audio Classification

Abstract

Mel-filterbanks are fixed, engineered audio features which emulate human perception and have been used through the history of audio understanding up to today. However, their undeniable qualities are counterbalanced by the fundamental limitations of handmade representations. In this work we show that we can train a single learnable frontend that outperforms mel-filterbanks on a wide range of audio signals, including speech, music, audio events and animal sounds, providing a general-purpose learned frontend for audio classification. To do so, we introduce a new principled, lightweight, fully learnable architecture that can be used as a drop-in replacement of mel-filterbanks. Our system learns all operations of audio features extraction, from filtering to pooling, compression and normalization, and can be integrated into any neural network at a negligible parameter cost. We perform multi-task training on eight diverse audio classification tasks, and show consistent improvements of our model over mel-filterbanks and previous learnable alternatives. Moreover, our system outperforms the current state-of-the-art learnable frontend on Audioset, with orders of magnitude fewer parameters.

Accepted at ICLR 2021

References in corpus (11)

• mixup: Beyond Empirical Risk Minimization
• SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
• Conv-TasNet: Surpassing Ideal Time-Frequency Magnitude Masking for Speech Separation
• Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition
• Deep Scattering Spectrum
• Automatic acoustic detection of birds through deep learning: the first Bird Audio Detection challenge
• End-to-end ASR: from Supervised to Semi-Supervised Learning with Modern Architectures
• Dual-path RNN: efficient long sequence modeling for time-domain single-channel speech separation
• Multiclass Language Identification using Deep Learning on Spectral Images of Audio Signals
• Reducing Bias in Production Speech Models
• Attention-based Wav2Text with Feature Transfer Learning

Cited by in corpus (7)

• Adaptive Representations of Sound for Automatic Insect Recognition
• Multimodal Self-Supervised Learning of General Audio Representations
• DCCRN+: Channel-wise Subband DCCRN with SNR Estimation for Speech Enhancement
• An Investigation of the Effectiveness of Phase for Audio Classification
• Free-Space Optical Spiking Neural Network
• Improving Sound Event Classification by Increasing Shift Invariance in Convolutional Neural Networks
• Complementing Handcrafted Features with Raw Waveform Using a Light-weight Auxiliary Model