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

arXiv: https://arxiv.org/abs/2103.00833 · HTML: https://mlsanity.com/paper/2103.00833/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 15 references that are in the map.
- First posted on arXiv 1 Mar 2021.

## Builds on (most cited first, 10 of 10)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Surrogate Gradient Learning in Spiking Neural Networks](https://mlsanity.com/paper/1901.09948/) (2019)
- [PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition](https://mlsanity.com/paper/1912.10211/) (2019)
- [Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1](https://mlsanity.com/paper/1602.02830/) (2016)
- [Sound Event Detection of Weakly Labelled Data with CNN-Transformer and Automatic Threshold Optimization](https://mlsanity.com/paper/1912.04761/) (2019)
- [Weakly Labelled AudioSet Tagging with Attention Neural Networks](https://mlsanity.com/paper/1903.00765/) (2019)
- [Technical report: supervised training of convolutional spiking neural networks with PyTorch](https://mlsanity.com/paper/1911.10124/) (2019)
- [Large-Scale Weakly Labeled Semi-Supervised Sound Event Detection in Domestic Environments](https://mlsanity.com/paper/1807.10501/) (2018)
- [Evaluation of post-processing algorithms for polyphonic sound event detection](https://mlsanity.com/paper/1906.06909/) (2019)

## Built on by (most cited first, 2 of 2)

- [Reliable Decision from Multiple Subtasks through Threshold Optimization: Content Moderation in the Wild](https://mlsanity.com/paper/2208.07522/) (2022)
- [Spiking neural networks trained via proxy](https://mlsanity.com/paper/2109.13208/) (2021)

## Abstract (arXiv)

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