ARTNet for Micro-Expression Recognition
Chao Wan, Wenbing Zhang, Yadong Chen, Liangliang Song, Peng Cheng · 2025
Read ARTNet for Micro-Expression Recognition by Chao Wan, Wenbing Zhang, Yadong Chen, Liangliang Song, Peng Cheng (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/pmc-12788191.txt.
Description
Micro-expressions are brief, faint facial movements that can reveal a person's true feelings, but they are hard for software to catch. Chao Wan and co-authors proposed a network for that task in 2025, called ARTNet. Most older methods track motion between video frames, yet the motion is often too weak and varies a lot from person to person.
ARTNet first boosts the motion gap between frames, making each expression stronger and easier to see. It then computes optical flow on the boosted frames and uses transformer layers to link features across them. Tests on three datasets confirmed that the approach works well, which the authors offer as a step toward more reliable emotion reading.
Catalog record
| CREATOR | Chao Wan, Wenbing Zhang, Yadong Chen, Liangliang Song, Peng Cheng |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |