Let’s connect at ECCV
Muhammad Osama Zeeshan
Computer Vision & Machine Learning Researcher
Montréal, Canada
Presenting CLIP-AUTT: Test-Time Personalization with Action Unit Prompting for Fine-Grained Video Emotion Recognition and organizing the ABAW 11th Ambivalence/Hesitancy Video Recognition Challenge at ECCV 2026.
About me
Computer vision and machine learning researcher with a PhD in Systems Engineering from École de technologie supérieure (ÉTS), Montréal. My research focuses on human-centered visual intelligence, multimodal learning, vision-language models, and personalized model adaptation under real-world distribution shifts.
ECCV 2026 Main Conference Paper
CLIP-AUTT: Test-Time Personalization with Action Unit Prompting for Fine-Grained Video Emotion Recognition
This work introduces an efficient Action-Unit-guided CLIP adaptation method for fine-grained video emotion recognition, focusing on interpretable facial dynamics and personalized video understanding.
ABAW 11th Challenge @ ECCV 2026
Part of the organizing team for the Ambivalence/Hesitancy (AH) Video Recognition Challenge at the 11th ABAW Workshop & Competition. The challenge focuses on video-level recognition of ambivalence and hesitation in real-world human behavior, using the BAH dataset.
Workshop publication
ABAW Workshop Paper @ ECCV 2026
BEACON: Behavior and Appearance Control for Subject-Specific Video Generation
This paper introduces a subject-specific video generation approach that controls both behavior and appearance, connecting personalized visual modeling with human-centered video understanding.
Stay in touch
Contact details
- osamaz.oz31@gmail.com
- Phone
- +1 438 357 6482
- Website
- osamazeeshan.github.io
- Muhammad Osama Zeeshan
- Google Scholar
- View profile