Cornell Personals Robotics: Videos

Anticipating Human Activities for Reactive Robotic Response

Hallucinating Humans for Robotic Scene Understanding

Descriptive Labelling of Activities

Learning to Place New Objects

PR2: Box packaging robot

Salad anyone? Robot slices cucumber

Contextually Guided Scene Understanding for RGBD data

Inferring articulated 3D structures

Scene Understanding using FeCCM

Learning to grasp new objects: dishwasher unloading

Response to Military Gestures

Links to videos

Grasping summary video
Grasping (Barrett 3-fingered hand) in cluttered environments. 1 (Cluttered), 2 (Kitchen), 3 (Cluttered), 4 (Cluttered), Simple scene - full sequence. (WMV)
Learning to Open New Doors. Full sequence, Montage: different new doors. (WMV)
Fetching a stapler from inside an office. MPEG-4 (30MB)
Fetching a stapler. Using a learned strategy for grasping novel objects, foveal-peripheral computer vision, depth perception and indoor navigation, and an MDP-based spoken dialog system, this shows STAIR fetching an item in response to a verbal request (6x speed): Quicktime (8MB), MPEG-4 (20MB), or Windows Media (16MB)
Unloading items from a dishwasher (2x speed): Windows Media (15MB).
Grasping previously unknown objects (3x speed): MPEG-4 (3MB), low resolution AVI (18MB), or high resolution AVI (146MB)