Response binning: Improved weak classifiers for boosting

Babak Rasolzadeh*, Lars Petersson, Niklas Pettersson

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

41 Citations (Scopus)

Abstract

This paper demonstrates the value of improving the discriminating strength of weak classifiers in the context of boosting by using response binning. The reasoning is centered around, but not limited to, the well known Haar-features used by Viola & Jones in their face detection/pedestrian detection systems. It is shown that using a weak classifier based on a single threshold is sub-optimal and in the case of the Haar-feature inadequate. A more general method for features with multi-modal responses is derived that is easily used in boosting mechanisms that accepts a confidence measure, such as the RealBoost algorithm. The method is evaluated by boosting a single stage classifier and compare the performance to previous approaches.

Original languageEnglish
Title of host publication2006 IEEE Intelligent Vehicles Symposium, IV 2006
Pages344-349
Number of pages6
Publication statusPublished - 2006
Externally publishedYes
Event2006 IEEE Intelligent Vehicles Symposium, IV 2006 - Meguro-Ku, Tokyo, Japan
Duration: 13 Jun 200615 Jun 2006

Publication series

NameIEEE Intelligent Vehicles Symposium, Proceedings

Conference

Conference2006 IEEE Intelligent Vehicles Symposium, IV 2006
Country/TerritoryJapan
CityMeguro-Ku, Tokyo
Period13/06/0615/06/06

Fingerprint

Dive into the research topics of 'Response binning: Improved weak classifiers for boosting'. Together they form a unique fingerprint.

Cite this