Occluded person re-identification with single-scale global representations
Occluded person re-identification (ReID) aims at re-identifying occluded pedestrians from occluded or holistic images taken across multiple cameras. Current state-of-the-art (SOTA) occluded ReID models rely on some auxiliary modules, including pose estimation, feature pyramid and graph matching modu...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2021
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/7313 https://ink.library.smu.edu.sg/context/sis_research/article/8316/viewcontent/Occluded_Person_Re_Identification_ICCV_2021_av_oa.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-8316 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-83162022-09-29T06:02:41Z Occluded person re-identification with single-scale global representations YAN, Cheng PANG, Guansong JIAO, Jile BAI, Xiao FENG, Xuetao SHEN, Chunhua Occluded person re-identification (ReID) aims at re-identifying occluded pedestrians from occluded or holistic images taken across multiple cameras. Current state-of-the-art (SOTA) occluded ReID models rely on some auxiliary modules, including pose estimation, feature pyramid and graph matching modules, to learn multi-scale and/or part-level features to tackle the occlusion challenges. This unfortunately leads to complex ReID models that (i) fail to generalize to challenging occlusions of diverse appearance, shape or size, and (ii) become ineffective in handling non-occluded pedestrians. However, real-world ReID applications typically have highly diverse occlusions and involve a hybrid of occluded and non-occluded pedestrians. To address these two issues, we introduce a novel ReID model that learns discriminative single-scale global-level pedestrian features by enforcing a novel exponentially sensitive yet bounded distance loss on occlusion-based augmented data. We show for the first time that learning single-scale global features without using these auxiliary modules is able to outperform the SOTA multi-scale and/or part-level feature-based models. Further, our simple model can achieve new SOTA performance in both occluded and non-occluded ReID, as shown by extensive results on three occluded and two general ReID benchmarks. Additionally, we create a large-scale occluded person ReID dataset with various occlusions in different scenes, which is significantly larger and contains more diverse occlusions and pedestrian dressings than existing occluded ReID datasets, providing a more faithful occluded ReID benchmark. The dataset is available at: https://git.io/OPReID 2021-10-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7313 info:doi/10.1109/ICCV48922.2021.01166 https://ink.library.smu.edu.sg/context/sis_research/article/8316/viewcontent/Occluded_Person_Re_Identification_ICCV_2021_av_oa.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Image and video retrieval Datasets and evaluation Representation learning Databases and Information Systems Numerical Analysis and Scientific Computing |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Image and video retrieval Datasets and evaluation Representation learning Databases and Information Systems Numerical Analysis and Scientific Computing |
spellingShingle |
Image and video retrieval Datasets and evaluation Representation learning Databases and Information Systems Numerical Analysis and Scientific Computing YAN, Cheng PANG, Guansong JIAO, Jile BAI, Xiao FENG, Xuetao SHEN, Chunhua Occluded person re-identification with single-scale global representations |
description |
Occluded person re-identification (ReID) aims at re-identifying occluded pedestrians from occluded or holistic images taken across multiple cameras. Current state-of-the-art (SOTA) occluded ReID models rely on some auxiliary modules, including pose estimation, feature pyramid and graph matching modules, to learn multi-scale and/or part-level features to tackle the occlusion challenges. This unfortunately leads to complex ReID models that (i) fail to generalize to challenging occlusions of diverse appearance, shape or size, and (ii) become ineffective in handling non-occluded pedestrians. However, real-world ReID applications typically have highly diverse occlusions and involve a hybrid of occluded and non-occluded pedestrians. To address these two issues, we introduce a novel ReID model that learns discriminative single-scale global-level pedestrian features by enforcing a novel exponentially sensitive yet bounded distance loss on occlusion-based augmented data. We show for the first time that learning single-scale global features without using these auxiliary modules is able to outperform the SOTA multi-scale and/or part-level feature-based models. Further, our simple model can achieve new SOTA performance in both occluded and non-occluded ReID, as shown by extensive results on three occluded and two general ReID benchmarks. Additionally, we create a large-scale occluded person ReID dataset with various occlusions in different scenes, which is significantly larger and contains more diverse occlusions and pedestrian dressings than existing occluded ReID datasets, providing a more faithful occluded ReID benchmark. The dataset is available at: https://git.io/OPReID |
format |
text |
author |
YAN, Cheng PANG, Guansong JIAO, Jile BAI, Xiao FENG, Xuetao SHEN, Chunhua |
author_facet |
YAN, Cheng PANG, Guansong JIAO, Jile BAI, Xiao FENG, Xuetao SHEN, Chunhua |
author_sort |
YAN, Cheng |
title |
Occluded person re-identification with single-scale global representations |
title_short |
Occluded person re-identification with single-scale global representations |
title_full |
Occluded person re-identification with single-scale global representations |
title_fullStr |
Occluded person re-identification with single-scale global representations |
title_full_unstemmed |
Occluded person re-identification with single-scale global representations |
title_sort |
occluded person re-identification with single-scale global representations |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2021 |
url |
https://ink.library.smu.edu.sg/sis_research/7313 https://ink.library.smu.edu.sg/context/sis_research/article/8316/viewcontent/Occluded_Person_Re_Identification_ICCV_2021_av_oa.pdf |
_version_ |
1770576309604319232 |