Scholarlink

<p>Recently, the advancement of deep learning has led to considerable breakthroughs in the automated detection of Action Units (AUs). Nevertheless, this field is still faced with various challenges, including limited subjects in shared datasets and the difficulty of collecting AU data as domain knowledge required for annotating AUs. These issues make it arduous for the model to generalize across different subjects and attain satisfactory performance on all AUs. To address these challenges, we propose two methods, namely AU-specific Querying (AQ) and Feedback Querying (FQ). AQ learns the global semantics of a particular AU, while FQ provides local structure information related to the AU semantic vector. The combination of these two operations enables the model to leverage both local features and global semantics of AUs. Moreover, Feedback Querying exhibits strong extensibility, which has led us to propose Cross-subject Querying (CQ). This method learns a subject-independent feature representation for each AU, resulting in improved generalization ability across different subjects. We demonstrate the effectiveness of our methods through visual presentations and ablation analysis. By combining all the strategies, our proposed AQ-CSL becomes the state-of-the-art model on the DISFA and BP4D datasets.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AQ-CSL: Attention Querying Facial Action Unit Detection Net with Cross Subject Learning

  • 期刊论文

  • 发表日期：

    2025年9月22日

查看全文
  • Wangding Zeng,
  • Yuanpeng Ji,
  • Donghao Chen,
  • Honggang Zhang

摘要

Recently, the advancement of deep learning has led to considerable breakthroughs in the automated detection of Action Units (AUs). Nevertheless, this field is still faced with various challenges, including limited subjects in shared datasets and the difficulty of collecting AU data as domain knowledge required for annotating AUs. These issues make it arduous for the model to generalize across different subjects and attain satisfactory performance on all AUs. To address these challenges, we propose two methods, namely AU-specific Querying (AQ) and Feedback Querying (FQ). AQ learns the global semantics of a particular AU, while FQ provides local structure information related to the AU semantic vector. The combination of these two operations enables the model to leverage both local features and global semantics of AUs. Moreover, Feedback Querying exhibits strong extensibility, which has led us to propose Cross-subject Querying (CQ). This method learns a subject-independent feature representation for each AU, resulting in improved generalization ability across different subjects. We demonstrate the effectiveness of our methods through visual presentations and ablation analysis. By combining all the strategies, our proposed AQ-CSL becomes the state-of-the-art model on the DISFA and BP4D datasets.