用生成式方法预测面部情绪，让模型学会表达不确定性和多解性。
AffectFlow-DINO: Uncertainty-Aware Multi-Task Affect Estimation via Conditional Rectified Flow
- 引入条件修正流架构，从单一预测变为不确定性感知的多解生成。
- 在11项任务中综合得分达1.177，比基线提升超过160%。
- 特别适合处理稀有表情类别，可不重新训练就大幅提升识别率。
我们提出AffectFlow-DINO，一个用于第11届ABAW挑战赛的多任务学习系统。该系统在冻结的DINOv3 ViT-S/16主干网络基础上，引入条件修正流解码头，以建模真实场景下面部行为的固有模糊性。不同于传统单一定值预测，模型学习条件生成分布，通过蒙特卡洛采样实现不确定性感知的多解输出。系统联合估计连续效价-唤醒度、分类八类面部表情、检测十二个动作单元。消融实验表明，修正流解码持续提升确定性预测性能，尤其在效价-唤醒度估计上（CCC-V提升0.058）。后处理阈值校准能有效恢复严重不平衡的罕见类别表现（如恐惧：3.8% → 33.1%），无需重训练。结合主干微调与流重调，最终模型取得$\mathbf{P_{MTL}=1.177}$，显著优于官方基线$P_{MTL}=0.45$。
原文摘要 · Abstract (English)
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 $P_{MTL}=0.45$.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。