# Redundancy Reduction Twins Network: A Training framework for Multi-output Emotion Regression

arXiv: https://arxiv.org/abs/2206.09142 · HTML: https://mlsanity.com/paper/2206.09142/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Jun 2022.

## Built on by (most cited first, 5 of 5)

- [Self-Supervised Attention Networks and Uncertainty Loss Weighting for Multi-Task Emotion Recognition on Vocal Bursts](https://mlsanity.com/paper/2209.07384/) (2022)
- [A Hierarchical Regression Chain Framework for Affective Vocal Burst Recognition](https://mlsanity.com/paper/2303.08027/) (2023)
- [Understanding Self-Supervised Learning of Speech Representation via Invariance and Redundancy Reduction](https://mlsanity.com/paper/2309.03619/) (2023)
- [Noise robust distillation of self-supervised speech models via correlation metrics](https://mlsanity.com/paper/2312.12153/) (2023)
- [The NeurIPS 2023 Machine Learning for Audio Workshop: Affective Audio Benchmarks and Novel Data](https://mlsanity.com/paper/2403.14048/) (2024)

## Abstract (arXiv)

In this paper, we propose the Redundancy Reduction Twins Network (RRTN), a redundancy reduction training framework that minimizes redundancy by measuring the cross-correlation matrix between the outputs of the same network fed with distorted versions of a sample and bringing it as close to the identity matrix as possible. RRTN also applies a new loss function, the Barlow Twins loss function, to help maximize the similarity of representations obtained from different distorted versions of a sample. However, as the distribution of losses can cause performance fluctuations in the network, we also propose the use of a Restrained Uncertainty Weight Loss (RUWL) or joint training to identify the best weights for the loss function. Our best approach on CNN14 with the proposed methodology obtains a CCC over emotion regression of 0.678 on the ExVo Multi-task dev set, a 4.8% increase over a vanilla CNN 14 CCC of 0.647, which achieves a significant difference at the 95% confidence interval (2-tailed).