Skip to yearly menu bar Skip to main content


COHESIV: Contrastive Object and Hand Embedding Segmentation In Video

Dandan Shan · Richard Higgins · David Fouhey

Keywords: [ Contrastive Learning ] [ Deep Learning ]


In this paper we learn to segment hands and hand-held objects from motion. Our system takes a single RGB image and hand location as input to segment the hand and hand-held object. For learning, we generate responsibility maps that show how well a hand's motion explains other pixels' motion in video. We use these responsibility maps as pseudo-labels to train a weakly-supervised neural network using an attention-based similarity loss and contrastive loss. Our system outperforms alternate methods, achieving good performance on the 100DOH, EPIC-KITCHENS, and HO3D datasets.

Chat is not available.