DRKF: Decoupled Representations with Knowledge Fusion for Multimodal Emotion Recognition 488 Conference: Proceedings of the 33rd ACM International Conference on Multimedia; October 27–31, 2025; Dublin, Ireland Proceedings of the 33rd ACM International Conference on Multimedia (MM ’25), October 27–31
Peiyuan Jiang, Yao Liu, Qiao Liu, Zongshun Zhang, Jiaye Yang, Lu Liu, Daibing Yao · 2025
Read DRKF: Decoupled Representations with Knowledge Fusion for Multimodal Emotion Recognition 488 Conference: Proceedings of the 33rd ACM International Conference on Multimedia; October 27–31, 2025; Dublin, Ireland Proceedings of the 33rd ACM International Conference on Multimedia (MM ’25), October 27–31 by Peiyuan Jiang, Yao Liu, Qiao Liu, Zongshun Zhang, Jiaye Yang, Lu Liu, Daibing Yao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-01644.txt.
Description
Peiyuan Jiang and six co-authors present a 2025 arXiv paper on reading emotion from voice, face, and text at once. Such systems often stumble because each channel carries feeling in its own way, and the channels can even disagree. A cheerful voice with a tense face is a hard case for a machine.
Their method, DRKF, first pulls apart what the channels share from what each one holds alone. A second part then picks the strongest channel and blends the rest into it. A guard part keeps signs of disagreement, so a wrong pick does not erase useful clues. Tests on three known data sets, IEMOCAP, MELD, and M3ED, gave the best scores so far. The team shares its code, so other groups can build on the design.
Catalog record
| CREATOR | Peiyuan Jiang, Yao Liu, Qiao Liu, Zongshun Zhang, Jiaye Yang, Lu Liu, Daibing Yao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-01644