Skip to Main content Skip to Navigation
Conference papers

Fast threshold optimization for multi-label audio tagging using Surrogate gradient learning

Abstract : Multi-label audio tagging consists of assigning sets of tags to audio recordings. At inference time, thresholds are applied on the confidence scores outputted by a probabilistic classifier, in order to decide which classes are detected active. In this work, we consider having at disposal a trained classifier and we seek to automatically optimize the decision thresholds according to a performance metric of interest, in our case F-measure (micro-F1). We propose a new method, called SGL-Thresh for Surrogate Gradient Learning of Thresholds, that makes use of gradient descent. Since F1 is not differentiable, we propose to approximate the thresholding operation gradients with the gradients of a sigmoid function. We report experiments on three datasets, using state-of-the-art pre-trained deep neural networks. In all cases, SGL-Thresh outperformed three other approaches: a default threshold value (defThresh), an heuristic search algorithm and a method estimating F1 gradients numerically. It reached 54.9\% F1 on AudioSet eval, compared to 50.7% with defThresh. SGL-Thresh is very fast and scalable to a large number of tags. To facilitate reproducibility, data and source code in Pytorch are available online: https://github.com/topel/SGL-Thresh
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03153644
Contributor : Thomas Pellegrini <>
Submitted on : Friday, February 26, 2021 - 3:38:35 PM
Last modification on : Thursday, March 18, 2021 - 2:16:12 PM

Identifiers

  • HAL Id : hal-03153644, version 1
  • ARXIV : 2103.00833

Citation

Thomas Pellegrini, Timothée Masquelier. Fast threshold optimization for multi-label audio tagging using Surrogate gradient learning. IEEE International Conference on Acoustics, Speech and Signal Processing, Jun 2021, Toronto, Canada. ⟨hal-03153644⟩

Share

Metrics

Record views

118

Files downloads

40