Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published 2004 | Published
Book Section - Chapter Open

An improved scheme for detection and labelling in Johansson displays

Abstract

Consider a number of moving points, where each point is attached to a joint of the human body and projected onto an image plane. Johannson showed that humans can effortlessly detect and recognize the presence of other humans from such displays. This is true even when some of the body points are missing (e.g. because of occlusion) and unrelated clutter points are added to the display. We are interested in replicating this ability in a machine. To this end, we present a labelling and detection scheme in a probabilistic framework. Our method is based on representing the joint probability density of positions and velocities of body points with a graphical model, and using Loopy Belief Propagation to calculate a likely interpretation of the scene. Furthermore, we introduce a global variable representing the body's centroid. Experiments on one motion-captured sequence suggest that our scheme improves on the accuracy of a previous approach based on triangulated graphical models, especially when very few parts are visible. The improvement is due both to the more general graph structure we use and, more significantly, to the introduction of the centroid variable.

Additional Information

We are very grateful to Max Welling, who Ørst proposed the idea of using LBP to solve for the optimal labelling in a 2001 Research Note, and who gave many useful suggestion. Sequences W1 and W2 used in the experiments were collected by L. Goncalves and E. di Bernando. This work was partially funded by the NSF Center for Neuromorphic Systems Engineering grant EEC-9402726 and by the ONR MURI grant N00014-01-1-0890.

Attached Files

Published - fanti_polito_perona_nips03.pdf

Files

fanti_polito_perona_nips03.pdf
Files (187.1 kB)
Name Size Download all
md5:ac604209000adf7dfdc912489c03a19a
187.1 kB Preview Download

Additional details

Created:
August 19, 2023
Modified:
January 13, 2024