# Learning from Annotation Uncertainty: Entropy-Aware Curriculum for Speech Emotion Recognition

arXiv: https://arxiv.org/abs/2606.27536 · HTML: https://mlsanity.com/paper/2606.27536/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 15 references that are in the map.
- First posted on arXiv 25 Jun 2026.

## Builds on (most cited first, 6 of 6)

- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [The MSP-Podcast Corpus](https://mlsanity.com/paper/2509.09791/) (2025)
- [End-to-End Label Uncertainty Modeling in Speech Emotion Recognition using Bayesian Neural Networks and Label Distribution Learning](https://mlsanity.com/paper/2209.15449/) (2022)
- [Modeling speech emotion with label variance and analyzing performance across speakers and unseen acoustic conditions](https://mlsanity.com/paper/2503.22711/) (2025)
- [Learning Annotation Consensus for Continuous Emotion Recognition](https://mlsanity.com/paper/2505.21196/) (2025)
- [Switchboard-Affect: Emotion Perception Labels from Conversational Speech](https://mlsanity.com/paper/2510.13906/) (2025)

## Abstract (arXiv)

Speech emotion recognition (SER) often relies on hard consensus labels that collapse annotator disagreement. We study distribution-based supervision for 9-class SER on MSP-Podcast 2.0 using a WavLM-Base multitask model for categorical emotion and dimensional VAD. Hard-label training is compared with targets from primary and merged primary--secondary annotator vote distributions. Distributional objectives improve alignment with human vote distributions, reducing JSD/KLD relative to hard-label training. Analysis shows that hard supervision partly benefits from assigning ambiguous utterances to the residual Other class, whereas distributional supervision redistributes uncertainty across emotion categories. Entropy-stratified evaluation shows that high-ambiguity utterances remain challenging, but distribution-based supervision better captures perceptual uncertainty. These findings support moving beyond hard labels toward targets that reflect listener disagreement.