Infused Suppression Of Magnification Artefacts For Micro-AU Detection

2025-04-12

AI Summary

Micro-expression AU detection faces challenges including subtle facial movements, short durations, and motion amplification artifacts—such as illumination inconsistencies and projection distortions—particularly undermining generalization in cross-database and multi-AU joint recognition tasks. To address these issues, we propose InfuseNet, a hierarchical unitary feature injection framework. Unlike conventional approaches relying on motion-amplified image reconstruction, InfuseNet directly leverages latent features generated during the amplification process. It introduces motion-contextual constraints to localize AU-responsive regions and applies unitary transformations to suppress feature distortion. The framework integrates optical flow modeling, latent-space feature distillation, and multi-database joint training. Under the CD6ME evaluation protocol, InfuseNet achieves significant improvements over state-of-the-art methods. Quantitative results demonstrate its effectiveness in mitigating amplification-induced artifacts, enhancing AU detection accuracy, and substantially improving cross-database generalization.

Technology Category

• Computer Vision: Motion & Tracking
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Affective Computing

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Facial micro-expressions are spontaneous, brief and subtle facial motions that unveil the underlying, suppressed emotions. Detecting Action Units (AUs) in micro-expressions is crucial because it yields a finer representation of facial motions than categorical emotions, effectively resolving the ambiguity among different expressions. One of the difficulties in micro-expression analysis is that facial motions are subtle and brief, thereby increasing the difficulty in correlating facial motion features to AU occurrence. To bridge the subtlety issue, flow-related features and motion magnification are a few common approaches as they can yield descriptive motion changes and increased motion amplitude respectively. While motion magnification can amplify the motion changes, it also accounts for illumination changes and projection errors during the amplification process, thereby creating motion artefacts that confuse the model to learn inauthentic magnified motion features. The problem is further aggravated in the context of a more complicated task where more AU classes are analyzed in cross-database settings. To address this issue, we propose InfuseNet, a layer-wise unitary feature infusion framework that leverages motion context to constrain the Action Unit (AU) learning within an informative facial movement region, thereby alleviating the influence of magnification artefacts. On top of that, we propose leveraging magnified latent features instead of reconstructing magnified samples to limit the distortion and artefacts caused by the projection inaccuracy in the motion reconstruction process. Via alleviating the magnification artefacts, InfuseNet has surpassed the state-of-the-art results in the CD6ME protocol. Further quantitative studies have also demonstrated the efficacy of motion artefacts alleviation.

Problem

• Detecting subtle facial micro-expressions for emotion analysis
• Reducing motion artefacts in magnified facial action units
• Improving cross-database AU classification accuracy

Innovation

• Layer-wise unitary feature infusion framework
• Leveraging magnified latent features
• Constraining AU learning within facial regions