Identifying exactly when an action (like "cutting") starts and ends.
Often includes synchronized gaze data (where the person is looking) Content and Activity
The file "g4_01136.mp4" is a technical video clip frequently used in computer vision research, specifically within the . This dataset is a cornerstone for studying human activity recognition and hand-object interactions from a first-person (egocentric) perspective. Overview and Context g4_01136.mp4
Modeling how a person’s eyes move toward an object before their hands touch it.
A consistent kitchen laboratory setup used across the "g4" (Group 4) subset of the data. Technical Significance Identifying exactly when an action (like "cutting") starts
Recognizing kitchen tools and ingredients from shifting, shaky angles.
🎥 This video is often cited in papers involving or Transformers designed for video understanding. It serves as a "real-world" challenge because of motion blur, hand occlusions, and the visual complexity of a cluttered kitchen. Overview and Context Modeling how a person’s eyes
High frequency of hand-to-object contact (e.g., opening jars, slicing vegetables, pouring liquids).