# hierarchical network with decoupled knowledge distillation for speech emotion recognition

arXiv: https://arxiv.org/abs/2303.05134 · HTML: https://mlsanity.com/paper/2303.05134/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal distillation, one of 12 directions in Knowledge distillation.
- Builds on 5 of its 26 references that are in the map.
- First posted on arXiv 9 Mar 2023.

## Placed in

- Knowledge distillation / [Multimodal distillation](https://mlsanity.com/distill/multimodal-distillation/)

## Builds on (most cited first, 5 of 5)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Decoupled Knowledge Distillation](https://mlsanity.com/paper/2203.08679/) (2022)
- [Show, Attend and Distill:Knowledge Distillation via Attention-based Feature Matching](https://mlsanity.com/paper/2102.02973/) (2021)
- [Speech Emotion Recognition with Co-Attention based Multi-level Acoustic Information](https://mlsanity.com/paper/2203.15326/) (2022)
- [Efficient Speech Emotion Recognition Using Multi-Scale CNN and Attention](https://mlsanity.com/paper/2106.04133/) (2021)

## Built on by (most cited first, 1 of 1)

- [Multi-Teacher Language-Aware Knowledge Distillation for Multilingual Speech Emotion Recognition](https://mlsanity.com/paper/2506.08717/) (2025)

## Abstract (arXiv)

The goal of Speech Emotion Recognition (SER) is to enable computers to recognize the emotion category of a given utterance in the same way that humans do. The accuracy of SER is strongly dependent on the validity of the utterance-level representation obtained by the model. Nevertheless, the ``dark knowledge" carried by non-target classes is always ignored by previous studies. In this paper, we propose a hierarchical network, called DKDFMH, which employs decoupled knowledge distillation in a deep convolutional neural network with a fused multi-head attention mechanism. Our approach applies logit distillation to obtain higher-level semantic features from different scales of attention sets and delve into the knowledge carried by non-target classes, thus guiding the model to focus more on the differences between sentiment features. To validate the effectiveness of our model, we conducted experiments on the Interactive Emotional Dyadic Motion Capture (IEMOCAP) dataset. We achieved competitive performance, with 79.1% weighted accuracy (WA) and 77.1% unweighted accuracy (UA). To the best of our knowledge, this is the first time since 2015 that logit distillation has been returned to state-of-the-art status.