Bridging the Gap for Test-Time Multimodal Sentiment Analysis
Zirun Guo, Tao Jin, Wenlong Xu, Wang Lin, Yangyang Wu · 2024
Read Bridging the Gap for Test-Time Multimodal Sentiment Analysis by Zirun Guo, Tao Jin, Wenlong Xu, Wang Lin, Yangyang Wu (2024) 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-2412-07121.txt.
Description
Zirun Guo, Tao Jin, Wenlong Xu, Wang Lin, and Yangyang Wu adapt sentiment models while they run in a 2024 computing paper. Multimodal sentiment analysis reads feeling from text, voice, and face together. Models trained once decay when the data stream shifts, as real streams do. Retraining usually needs the original data, which may be private or huge. Test-time adaptation fixes the model on the fly, using only what it sees now.
Their two strategies enforce consistency across modalities and mint cautious pseudo-labels to learn from. Across experiments the pair lifted performance steadily. Adapting at run time suits live settings like monitoring interviews or calls, where yesterday's data no longer matches today's speakers.
Catalog record
| CREATOR | Zirun Guo, Tao Jin, Wenlong Xu, Wang Lin, Yangyang Wu |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-07121