NTHRYS › Internships › Affective Science › Facial Action Unit Detection & Coding

Adversarial Robustness of Deep Learning Action Unit Detection Systems

Affective Science

Facial Action Unit Detection & Coding

Variant

Pay · Join

Step 4 of 5Select variant

Field

▾

Category

▾

Focused area

▾

You might prefer

Deep Convolutional Networks for Micro-Expression Recognition Cross-Cultural Facial Action Unit Variation and Coding Standards Real-Time 3D Facial Geometry and Action Unit Coupling Analysis Temporal Dynamics of Action Unit Onset, Apex, and Offset Phases Multimodal Integration of Facial Actions with Vocal Prosody Features Action Unit Prediction from Brain Activity Using fMRI Neuroimaging Data Spontaneous versus Posed Facial Action Unit Kinematic Discrimination Facial Action Unit Compound Effects and Co-Occurrence Modeling Domain Adaptation Techniques for Cross-Database Action Unit Transfer Learning

Affective Science › Facial Action Unit Detection & Coding

Adversarial Robustness of Deep Learning Action Unit Detection Systems

This research systematically evaluates vulnerabilities of neural networks to adversarial perturbations designed to fool action unit detection systems under realistic conditions. The investigation identifies critical robustness limitations and develops defensive mechanisms essential for deploying facial analysis systems in high-stakes applications.

🎓 TYPE

🌐 MODE

⏱

📚 Academic: Thesis & PPT assistance included🧪 Tech: Master the protocols hands-on📝 Research > 3 months: Publication co-authorship in a Scopus-indexed journal

Select your preferenceChoose Type, Mode, Duration to view Titles

Internship Type

Mode

Duration

🎯

Choose your preferences above

Select Type, Mode and Duration to view available internship titles and fees.