APP logo

Activity Perception Project


Introduction

Vision Research Group
Artificial Intelligence Laboratory (CSAIL)
Massachusetts Institute of Technology


Introduction


Research Overview


Area Map


Camera views


Mosaic views


Analysis of Activities


People

Projects

Publications

Links

Internal

(for AI Lab access only)

Welcome to the APP

We are developing computer-vision and machine-learning techniques for passive observation of activities in natural settings using a distributed camera network. The visual data thus gathered is used for object recognition/classification, classification of activities of participants in the scene, and the detection of regular patterns of usage and unusual events in the site.

By following the links on the left, you will find descriptions of our object-detection/tracking infrastructure and current and past research on activity perception.



Area Map: aerial views of the site.

Cameras: views from the actual cameras used.

Mosaics: shows areas covered by groups of cameras, in the form of panoramic mosaics.

Analysis of Activities: a list of capabilities of the project.

Aerial view
Aerial view of Tech Square area, where our lab is located
Fields-of-view of cameras (in green)






Last updated 05 August 2003.  Questions/comments? E-mail app-web at ai.mit.edu