1 paper across 1 session
Regularize the model attention in a transformer (VLM) using human gaze to enhance egocentric activity prediction and understanding.
