AffectFlow-DINO: Uncertainty-Aware Multi-Task Affect Estimation via Conditional Rectified Flow
- 类型:arxiv
- 标识:2607.13250
- 链接:https://arxiv.org/abs/2607.13250
- 主分类:multimodal
- 形态:method
- 被引:0
- 被引来源:Semantic Scholar + OpenAlex
- S2被引:0
- OpenAlex被引:0
- 影响力被引:0
- TLDR: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, enabling uncertainty-aware one-to-many predictions through Monte Carlo sampling.
- OpenAlex ID:W7168709147
- OpenAlex DOI:10.48550/arxiv.2607.13250
- DOI:10.48550/arxiv.2607.13250
- DOI来源:OpenAlex
- 开放获取:green
- 开放获取链接:https://doi.org/10.48550/arxiv.2607.13250
- OpenAlex更新:2026-07-19
- 待LLM分类:否
- 标题中文:AffectFlow-DINO:基于条件 Rectified Flow 的不确定性感知多任务情感估计
- TLDR中文:面向第 11 届 ABAW 挑战赛的多任务学习系统,在标准确定性架构基础上扩展条件 Rectified Flow 头,建模真实场景下面部行为固有的模糊性,借助蒙特卡洛采样实现不确定性感知的一对多预测。
- 成熟度:research
- 场景:affect-estimation、facial-behavior、uncertainty
- 来源文件:
- /inbox/tom/_candidates/2026-07-17-agent-rag-longcontext-candidates.json
- [S2 enrich]
- [OpenAlex backfill]