DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING

Yan Bai; Jie Lin; Vijay Chandrasekhar; Yihang Lou; Shiqi Wang; Ling-Yu Duan; Tiejun Huang; Alex Kot

doi:10.1109/ICIP.2017.8296307

DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING

Yan Bai
, Jie Lin
, Vijay Chandrasekhar
, Yihang Lou
, Shiqi Wang
, Ling-Yu Duan^*
, Tiejun Huang
, Alex Kot

^*Corresponding author for this work

Department of Computer Science

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

Abstract

In this work, we study the problem of deep global descriptors for video matching with regional feature pooling. We aim to analyze the joint effect of ROI (Region of Interest) size and pooling moment on video matching performance. To this end, we propose to mathematically model the distribution of video matching function with a pooling function nested in. Matching performance can be estimated by the separability of these class-conditional distributions between matching and non-matching pairs. Empirical studies on the challenging MPEG CDVA dataset demonstrate that performance trends are consistent with the estimation and experimental results, though the theoretical model is largely simplified compared to video matching and retrieval in practice.

Original language	English
Title of host publication	2017 IEEE International Conference on Image Processing, ICIP 2017 - Proceedings
Publisher	IEEE
Pages	380-384
ISBN (Electronic)	9781509021758
DOIs	https://doi.org/10.1109/ICIP.2017.8296307
Publication status	Published - Sept 2017
Event	24th IEEE International Conference on Image Processing (ICIP 2017) - China National Convention Center, Beijing, China Duration: 17 Sept 2017 → 20 Sept 2017 http://2017.ieeeicip.org/ https://ieeexplore.ieee.org/xpl/conhome/8267582/proceeding

Publication series

Name	IEEE International Conference on Image Processing, ICIP
Publisher	IEEE
ISSN (Print)	1522-4880
ISSN (Electronic)	2381-8549

Conference

Conference	24th IEEE International Conference on Image Processing (ICIP 2017)
Abbreviated title	ICIP
Place	China
City	Beijing
Period	17/09/17 → 20/09/17
Internet address	http://2017.ieeeicip.org/ https://ieeexplore.ieee.org/xpl/conhome/8267582/proceeding

Research Keywords

Convolutional Neural Networks
Global Descriptor
Pooling
Video Matching
Video Retrieval

Access Output

10.1109/ICIP.2017.8296307

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{a10167d6fe2942b8b11baf195e71a176,

title = "DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING",

abstract = "In this work, we study the problem of deep global descriptors for video matching with regional feature pooling. We aim to analyze the joint effect of ROI (Region of Interest) size and pooling moment on video matching performance. To this end, we propose to mathematically model the distribution of video matching function with a pooling function nested in. Matching performance can be estimated by the separability of these class-conditional distributions between matching and non-matching pairs. Empirical studies on the challenging MPEG CDVA dataset demonstrate that performance trends are consistent with the estimation and experimental results, though the theoretical model is largely simplified compared to video matching and retrieval in practice.",

keywords = "Convolutional Neural Networks, Global Descriptor, Pooling, Video Matching, Video Retrieval",

author = "Yan Bai and Jie Lin and Vijay Chandrasekhar and Yihang Lou and Shiqi Wang and Ling-Yu Duan and Tiejun Huang and Alex Kot",

year = "2017",

month = sep,

doi = "10.1109/ICIP.2017.8296307",

language = "English",

series = "IEEE International Conference on Image Processing, ICIP",

publisher = "IEEE",

pages = "380--384",

booktitle = "2017 IEEE International Conference on Image Processing, ICIP 2017 - Proceedings",

address = "United States",

note = "24th IEEE International Conference on Image Processing (ICIP 2017), ICIP ; Conference date: 17-09-2017 Through 20-09-2017",

url = "http://2017.ieeeicip.org/, https://ieeexplore.ieee.org/xpl/conhome/8267582/proceeding",

}

Bai, Y, Lin, J, Chandrasekhar, V, Lou, Y, Wang, S, Duan, L-Y, Huang, T & Kot, A 2017, DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING. in 2017 IEEE International Conference on Image Processing, ICIP 2017 - Proceedings. IEEE International Conference on Image Processing, ICIP, IEEE, pp. 380-384, 24th IEEE International Conference on Image Processing (ICIP 2017), Beijing, China, 17/09/17. https://doi.org/10.1109/ICIP.2017.8296307

DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING. / Bai, Yan; Lin, Jie; Chandrasekhar, Vijay et al.
2017 IEEE International Conference on Image Processing, ICIP 2017 - Proceedings. IEEE, 2017. p. 380-384 (IEEE International Conference on Image Processing, ICIP).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING

AU - Bai, Yan

AU - Lin, Jie

AU - Chandrasekhar, Vijay

AU - Lou, Yihang

AU - Wang, Shiqi

AU - Duan, Ling-Yu

AU - Huang, Tiejun

AU - Kot, Alex

PY - 2017/9

Y1 - 2017/9

N2 - In this work, we study the problem of deep global descriptors for video matching with regional feature pooling. We aim to analyze the joint effect of ROI (Region of Interest) size and pooling moment on video matching performance. To this end, we propose to mathematically model the distribution of video matching function with a pooling function nested in. Matching performance can be estimated by the separability of these class-conditional distributions between matching and non-matching pairs. Empirical studies on the challenging MPEG CDVA dataset demonstrate that performance trends are consistent with the estimation and experimental results, though the theoretical model is largely simplified compared to video matching and retrieval in practice.

AB - In this work, we study the problem of deep global descriptors for video matching with regional feature pooling. We aim to analyze the joint effect of ROI (Region of Interest) size and pooling moment on video matching performance. To this end, we propose to mathematically model the distribution of video matching function with a pooling function nested in. Matching performance can be estimated by the separability of these class-conditional distributions between matching and non-matching pairs. Empirical studies on the challenging MPEG CDVA dataset demonstrate that performance trends are consistent with the estimation and experimental results, though the theoretical model is largely simplified compared to video matching and retrieval in practice.

KW - Convolutional Neural Networks

KW - Global Descriptor

KW - Pooling

KW - Video Matching

KW - Video Retrieval

UR - https://www.scopus.com/pages/publications/85045295403

UR - https://www.scopus.com/pages/publications/85045295403#tab=impact

U2 - 10.1109/ICIP.2017.8296307

DO - 10.1109/ICIP.2017.8296307

M3 - RGC 32 - Refereed conference paper (with host publication)

T3 - IEEE International Conference on Image Processing, ICIP

SP - 380

EP - 384

BT - 2017 IEEE International Conference on Image Processing, ICIP 2017 - Proceedings

PB - IEEE

T2 - 24th IEEE International Conference on Image Processing (ICIP 2017)

Y2 - 17 September 2017 through 20 September 2017

ER -

DEEP REGIONAL FEATURE POOLING FOR VIDEO MATCHING

Abstract

Publication series

Conference

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this