Learning to count objects in images

Victor Lempitsky, Andrew Zisserman

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

790 Citations (Scopus)

Abstract

We propose a new supervised learning framework for visual object counting tasks, such as estimating the number of cells in a microscopic image or the number of humans in surveillance video frames. We focus on the practically-attractive case when the training images are annotated with dots (one dot per object). Our goal is to accurately estimate the count. However, we evade the hard task of learning to detect and localize individual object instances. Instead, we cast the problem as that of estimating an image density whose integral over any image region gives the count of objects within that region. Learning to infer such density can be formulated as a minimization of a regularized risk quadratic cost function. We introduce a new loss function, which is well-suited for such learning, and at the same time can be computed efficiently via a maximum subarray algorithm. The learning can then be posed as a convex quadratic program solvable with cutting-plane optimization. The proposed framework is very flexible as it can accept any domain-specific visual features. Once trained, our system provides accurate object counts and requires a very small time overhead over the feature extraction step, making it a good candidate for applications involving real-time processing or dealing with huge amount of visual data.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 23
Subtitle of host publication24th Annual Conference on Neural Information Processing Systems 2010, NIPS 2010
Publication statusPublished - 2010
Externally publishedYes
Event24th Annual Conference on Neural Information Processing Systems 2010, NIPS 2010 - Vancouver, BC, Canada
Duration: 6 Dec 20109 Dec 2010

Publication series

NameAdvances in Neural Information Processing Systems 23: 24th Annual Conference on Neural Information Processing Systems 2010, NIPS 2010

Conference

Conference24th Annual Conference on Neural Information Processing Systems 2010, NIPS 2010
Country/TerritoryCanada
CityVancouver, BC
Period6/12/109/12/10

Fingerprint

Dive into the research topics of 'Learning to count objects in images'. Together they form a unique fingerprint.

Cite this