AffectFlow-DINO: Uncertainty-Aware Multi-Task Affect Estimation via Conditional Rectified Flow

Researchers introduced AffectFlow-DINO, a multi-task learning system for the 11th ABAW challenge that uses conditional rectified flow to model uncertainty in facial affect estimation. The system jointly estimates valence-arousal, classifies facial expressions, and detects action units, achieving a multi-task learning score of 1.177, significantly outperforming the official baseline of 0.45.

arXiv:2607.13250v1 Announce Type: new Abstract: We present \textbf{AffectFlow-DINO}, a multi-task learning system for the 11th ABAW challenge that extends a standard deterministic architecture with a conditional rectified-flow head to model the inherent ambiguity of in-the-wild facial behavior. Instead of predicting a single affect estimate, the model learns a conditional generative distribution, enabling uncertainty-aware one-to-many predictions through Monte Carlo sampling. The system jointly estimates continuous valence-arousal, classifies eight facial expressions, and detects twelve Action Units from static face images. Built on a frozen DINOv3 ViT-S/16 backbone, extensive ablation studies show that rectified-flow decoding consistently improves deterministic prediction, particularly for valence-arousal estimation CCC-V $+0.058$ . We further show that post-hoc threshold calibration effectively recovers performance on severely imbalanced rare classes e.g., Fear: $3.8\% \rightarrow 33.1\%$ without retraining. Combined with backbone fine-tuning and flow retuning, the final model achieves $\mathbf{P {MTL}=1.177}$, substantially outperforming the official challenge baseline of $P {MTL}=0.45$.