Disentangling Gaze Control
Shalini discusses the challenges of using unlabeled video data to control head and gaze movements independently. She highlights the complexities introduced by varying lighting conditions during recordings and the innovative self-learning techniques developed to disentangle these factors. The process not only enhances gaze tracking but also ensures consistent lighting effects, making the research both intriguing and practical.In this clip
From this podcast
The AI Podcast
NVIDIA’s Shalini De Mello Talks Self-Supervised AI, NeurIPS Successes - Ep. 140
Related Questions