EgoTV

EgoTV dataset consists of (task description, video) pairs with positive on negative task verification labels. By combining the six sub-tasks heat, clean, slice, cool, put, pick with different ordering constraints, there are 82 tasks for EgoTV. Tasks are instantiated with 130 target objects (excluding visual variations in shape, texture, and color) and 24 receptacle objects, totaling 1038 task object combinations. These are performed in 30 different kitchen scenes.

Source: EgoTV: Egocentric Task Verification from Natural Language Task Descriptions