Papers › Representation Learning and Identity Adversarial Training for Facial Behavior Understanding
Representation Learning and Identity Adversarial Training for Facial Behavior Understanding
Mang Ning, Albert Ali Salah, Itir Onal Ertugrul
Facial Action Unit (AU) detection has gained significant attention as it enables the breakdown of complex facial expressions into individual muscle movements. In this paper, we revisit two fundamental factors in AU detection: diverse and large-scale data and subject identity regularization. Motivated by recent advances in foundation models, we highlight the importance of data and introduce Face9M, a diverse dataset comprising 9 million facial images from multiple public sources. Pretraining a masked autoencoder on Face9M yields strong performance in AU detection and facial expression tasks. More importantly, we emphasize that the Identity Adversarial Training (IAT) has not been well explored in AU tasks. To fill this gap, we first show that subject identity in AU datasets creates shortcut learning for the model and leads to sub-optimal solutions to AU predictions. Secondly, we demonstrate that strong IAT regularization is necessary to learn identity-invariant features. Finally, we elucidate the design space of IAT and empirically show that IAT circumvents the identity-based shortcut learning and results in a better solution. Our proposed methods, Facial Masked Autoencoder (FMAE) and IAT, are simple, generic and effective. Remarkably, the proposed FMAE-IAT approach achieves new state-of-the-art F1 scores on BP4D (67.1\%), BP4D+ (66.8\%), and DISFA (70.1\%) databases, significantly outperforming previous work. We release the code and model at https://github.com/forever208/FMAE-IAT.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Facial Action Unit Detection | BP4D | FMAE-IAT | Average F1 | 67.1 | #1 of 10 | Archive leaderboard | report |
| Facial Action Unit Detection | BP4D | FMAE | Average F1 | 66.6 | #2 of 10 | Archive leaderboard | report |
| Facial Action Unit Detection | BP4D+ | FMAE-IAT | Average F1 | 66.8 | #1 of 3 | Archive leaderboard | report |
| Facial Action Unit Detection | BP4D+ | FMAE | Average F1 | 66.2 | #3 of 3 | Archive leaderboard | report |
| Facial Action Unit Detection | DISFA | FMAE_IAT | Average F1 | 70.1 | #2 of 8 | Archive leaderboard | report |
| Facial Action Unit Detection | DISFA | FMAE | Average F1 | 68.7 | #3 of 8 | Archive leaderboard | report |
| Facial Expression Recognition (FER) | AffectNet | FMAE | Accuracy (8 emotion) | 64.79 | #3 of 50 | Archive leaderboard | report |
| Facial Expression Recognition (FER) | RAF-DB | FMAE | Overall Accuracy | 93.45 | #2 of 35 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections