Training a multi-exit cascade with linear asymmetric classification for efficient object detection
-
Altmetric Citations
Wang, Peng; Shen, Chunhua; Zheng, Hong; Ren, Zhang
Description
Efficient visual object detection is of central interest in computer vision and pattern recognition due to its wide ranges of applications. Viola and Jones'detector has become a de facto framework [1]. In this work, we propose a new method to design a cascade of boosted classifiers for fast object detection, which combines linear asymmetric classification (LAC) into the recent multi-exit cascade structure. Therefore, the proposed method takes advantages of both LAC and the multi-exit cascade....[Show more]
dc.contributor.author | Wang, Peng | |
---|---|---|
dc.contributor.author | Shen, Chunhua | |
dc.contributor.author | Zheng, Hong | |
dc.contributor.author | Ren, Zhang | |
dc.coverage.spatial | Hong Kong China | |
dc.date.accessioned | 2015-12-10T23:04:36Z | |
dc.date.created | September 26-29 2010 | |
dc.identifier.uri | http://hdl.handle.net/1885/62434 | |
dc.description.abstract | Efficient visual object detection is of central interest in computer vision and pattern recognition due to its wide ranges of applications. Viola and Jones'detector has become a de facto framework [1]. In this work, we propose a new method to design a cascade of boosted classifiers for fast object detection, which combines linear asymmetric classification (LAC) into the recent multi-exit cascade structure. Therefore, the proposed method takes advantages of both LAC and the multi-exit cascade. Namely, (1) the multi-exit cascade structure collects all the scores of prior nodes for decision making at the current node, which reduces the loss of decision information; (2) LAC considers the asymmetric nature of the node training. We also show that the multi-exit cascade better meets the assumption of LAC learning than the standard Viola-Jones'cascade, both theoretically and empirically. Experiments confirm that our method outperforms existing methods such as Viola and Jones [1] and Wu et al. [2] on the MIT+CMU test data set. | |
dc.publisher | IEEE Signal Processing Society | |
dc.relation.ispartofseries | IEEE International Conference on Image Processing 2010 | |
dc.source | Proceedings of IEEE International Conference on Image Processing 2010 | |
dc.subject | Keywords: Boosted classifiers; Boosting; Cascade classifiers; Cascade structures; De facto; Decision information; Existing method; Face Detection; Linear asymmetric classifier; Object Detection; Test data; Visual objects; Computer vision; Decision making; Imaging s Boosting; Cascade classifier; Face detection; Linear asymmetric classifier | |
dc.title | Training a multi-exit cascade with linear asymmetric classification for efficient object detection | |
dc.type | Conference paper | |
local.description.notes | Imported from ARIES | |
local.description.refereed | Yes | |
dc.date.issued | 2010 | |
local.identifier.absfor | 080106 - Image Processing | |
local.identifier.ariespublication | u4334215xPUB698 | |
local.type.status | Published Version | |
local.contributor.affiliation | Wang, Peng, Beihang University | |
local.contributor.affiliation | Shen, Chunhua, College of Engineering and Computer Science, ANU | |
local.contributor.affiliation | Zheng, Hong, Beihang University | |
local.contributor.affiliation | Ren, Zhang, Beihang University | |
local.description.embargo | 2037-12-31 | |
local.bibliographicCitation.startpage | 61 | |
local.bibliographicCitation.lastpage | 64 | |
local.identifier.doi | 10.1109/ICIP.2010.5651599 | |
local.identifier.absseo | 970109 - Expanding Knowledge in Engineering | |
dc.date.updated | 2016-02-24T11:02:44Z | |
local.identifier.scopusID | 2-s2.0-78651069313 | |
Collections | ANU Research Publications |
Download
File | Description | Size | Format | Image |
---|---|---|---|---|
01_Wang_Training_a_multi-exit_cascade_2010.pdf | 441.2 kB | Adobe PDF | Request a copy |
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.
Updated: 17 November 2022/ Responsible Officer: University Librarian/ Page Contact: Library Systems & Web Coordinator