Anti-faces for detection

Daniel Keren, Margarita Osadchy, Craig Gotsman

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations


This paper offers a novel detection method, which works well even in the case of a complicated image collection -for instance, a frontal face under a large class of linear transformations. It was also successfully applied to detect 3D objects under different views. Call the class of images, which should be detected, a multi-template. The detection problem is solved by sequentially applying very simple filters (or detectors), which are designed to yield small results on the multi-template (hence \anti-faces"), and large results on \random" natural images. This is achieved by making use of a simple probabilistic assumption on the distribution of natural images, which is borne out well in practice, and by using a simple implicit representation of the multi-template. Only images which passed the threshold test imposed by the first detector are examined by the second detector, etc. The detectors have the added bonus that they act independently, so that their false alarms are uncorrelated; this results in a percentage of false alarms which exponentially decreases in the number of detectors. This, in turn, leads to a very fast detection algorithm, usually requiring (1 + δ)N operations to classify an N-pixel image, where δ < 0:5. Also, the algorithm requires no training loop. The suggested algorithm's performance favorably compares to the well- known eigenface and support vector machine based algorithms, and it is substantially faster.

Original languageEnglish (US)
Title of host publicationComputer Vision - ECCV 2000 - 6th European Conference on Computer Vision, Proceedings
EditorsDavid Vernon
PublisherSpringer Verlag
Number of pages15
ISBN (Print)3540676856
StatePublished - 2000
Externally publishedYes
Event6th European Conference on Computer Vision, ECCV 2000 - Dublin, Ireland
Duration: Jun 26 2000Jul 1 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other6th European Conference on Computer Vision, ECCV 2000

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Anti-faces for detection'. Together they form a unique fingerprint.

Cite this