Presentation 2010-06-10
GPU Implementation of CoHOG-based Pedestrian Detection
Hiroki SUGANO, Ryusuke MIYAMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Pedestrian detection from visual images, which is used by driver assistance or video surveillance, is a recent challenging problem. Co-occurrence histograms of oriented gradients (CoHOG) is a powerful feature descriptor for pedestrian detection and achieves the highest detection accuracy. However, its calculation cost is too large to calculate the CoHOG feature descriptor in real-time on state-of-the-art processors. In this paper, we focus on NVIDIA GPUs and propose optimized implementation of CoHOG-based pedestrian detection on them. Our proposed implementation does not depend on the number of processors in the GPU and has large scalability.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) CoHOG-based pedestrian detection / NVIDIA GPU / Parallel implementation / Real-time processing
Paper # SIS2010-11
Date of Issue

Conference Information
Committee SIS
Conference Date 2010/6/3(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Smart Info-Media Systems (SIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) GPU Implementation of CoHOG-based Pedestrian Detection
Sub Title (in English)
Keyword(1) CoHOG-based pedestrian detection
Keyword(2) NVIDIA GPU
Keyword(3) Parallel implementation
Keyword(4) Real-time processing
1st Author's Name Hiroki SUGANO
1st Author's Affiliation Dept. of Information Systems, Graduate School of Information Science, Nara Institute of Science and Technology()
2nd Author's Name Ryusuke MIYAMOTO
2nd Author's Affiliation Dept. of Information Systems, Graduate School of Information Science, Nara Institute of Science and Technology
Date 2010-06-10
Paper # SIS2010-11
Volume (vol) vol.110
Number (no) 74
Page pp.pp.-
#Pages 6
Date of Issue