Mar 5, 2021 · Abstract:We propose a two-stream convolutional network for audio recognition, that operates on time-frequency spectrogram inputs.

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 ...

Implementation of "Slow-Fast Auditory Streams for Audio Recognition, ICASSP, 2021" in PyTorch - ekazakos/auditory-slow-fast.

SLOW-FAST AUDITORY STREAMS FOR AUDIO RECOGNITION. International Conference on Acoustics, Speech and Signal Processing (ICASSP). Top-1 Accuracy (%). Split Model.

ABSTRACT. We propose a two-stream convolutional network for audio recognition, that operates on time-frequency spectrogram inputs.

This work learns Slow-Fast auditory streams with separable convolutions and multi-level lateral connections in a two-stream convolutional network for audio ...

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 ...

Mar 5, 2021 · We propose a two-stream convolutional network for audio recognition, that operates on time-frequency spectrogram inputs.

We train and evaluate the Auditory SlowFast (ASF) [27] and Self-Supervised Audio Spectrogram Transformer (SSAST) [26] audio encoder networks, with both a linear ...

EPIC-SOUNDS is a large scale dataset of audio annotations capturing temporal extents and class labels within the audio stream of the egocentric videos.