EmotionRankCLAP: Bridging Natural Language Speaking Styles and Ordinal Speech Emotion via Rank-N-Contrast
Shreeram Suresh, Lucas, Junchen, Carlos, Berrak · 2025
Read EmotionRankCLAP: Bridging Natural Language Speaking Styles and Ordinal Speech Emotion via Rank-N-Contrast by Shreeram Suresh, Lucas, Junchen, Carlos, Berrak (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-2505-23732.txt.
Description
Shreeram Suresh Chandra and co-authors teach a computer to rank feelings in speech rather than just name them, in a 2025 paper. Emotions come in degrees: one voice sounds a bit annoyed, another furious. Systems that match speech clips to text labels usually ignore that ordering, so near feelings blur together. The team brings in two standard feeling scales, pleasantness and energy, to sort the clips.
Their method, EmotionRankCLAP, trains speech and text descriptions together with a ranking-aware goal. Clips are pulled toward or pushed from each other based on their order along the feeling scales, not only on matching labels. The model also closes the gap between how speech and text are represented inside the system. In cross tests, asking speech to find its text match and back, the method beat earlier emotion systems at respecting the true order of feelings.
Catalog record
| CREATOR | Shreeram Suresh, Lucas, Junchen, Carlos, Berrak |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-23732