English

Slow-Fast Auditory Streams For Audio Recognition

Sound 2021-03-08 v1 Computer Vision and Pattern Recognition Audio and Speech Processing

Abstract

We propose a two-stream convolutional network for audio recognition, that operates on time-frequency spectrogram inputs. Following similar success in visual recognition, we learn Slow-Fast auditory streams with separable convolutions and multi-level lateral connections. The Slow pathway has high channel capacity while the Fast pathway operates at a fine-grained temporal resolution. We showcase the importance of our two-stream proposal on two diverse datasets: VGG-Sound and EPIC-KITCHENS-100, and achieve state-of-the-art results on both.

Keywords

Cite

@article{arxiv.2103.03516,
  title  = {Slow-Fast Auditory Streams For Audio Recognition},
  author = {Evangelos Kazakos and Arsha Nagrani and Andrew Zisserman and Dima Damen},
  journal= {arXiv preprint arXiv:2103.03516},
  year   = {2021}
}

Comments

Accepted for presentation at ICASSP 2021

R2 v1 2026-06-23T23:47:28.043Z