A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification
Person re-identification problems usually suffer from large subject appearance variations and limited training data. This paper proposes a novel physically motivated Color/Illuminance-Aware data-augmentation (CIADA) scheme and a style-adaptive fusion approach to address these issues. The CIADA schem...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9497884/ |
_version_ | 1818881354912235520 |
---|---|
author | Zhouchi Lin Chenyang Liu Wenbo Qi S. C. Chan |
author_facet | Zhouchi Lin Chenyang Liu Wenbo Qi S. C. Chan |
author_sort | Zhouchi Lin |
collection | DOAJ |
description | Person re-identification problems usually suffer from large subject appearance variations and limited training data. This paper proposes a novel physically motivated Color/Illuminance-Aware data-augmentation (CIADA) scheme and a style-adaptive fusion approach to address these issues. The CIADA scheme estimates the color/illuminance distribution from the training data via manifold learning and generates new samples under different color/illuminance perturbations to better capture objects’ appearance for mitigating the small-sample-size and color variation problems. A Color/Illuminance Aware Feature Augmentation (CIAFA) approach, which is applicable to state-of-the-art features and metric learning algorithms, is then proposed to integrate the features generated by the augmented samples for metric learning. A new Color/Illuminance-Aware Style Fusion (CIASF) scheme, which allows the learning and matching process to be performed independently on each pair of datasets generated for estimating a set of ‘local’ distance functions, is also proposed. A canonical correlation analysis-based weighting scheme is developed to fuse these local distances to an overall distance for recognition. This reduces the memory requirement and complexity over the original CIAFA. Experiments on common datasets show that the proposed methodologies substantially improve the performance of state-of-the-art subspace learning algorithms. It is applicable to both small and large datasets with hand-craft and deep features. |
first_indexed | 2024-12-19T15:00:32Z |
format | Article |
id | doaj.art-cec102f0912d4a6b87fe1ce9811089b3 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-12-19T15:00:32Z |
publishDate | 2021-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-cec102f0912d4a6b87fe1ce9811089b32022-12-21T20:16:34ZengIEEEIEEE Access2169-35362021-01-01911582611583810.1109/ACCESS.2021.31005719497884A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-IdentificationZhouchi Lin0Chenyang Liu1https://orcid.org/0000-0001-9997-4830Wenbo Qi2https://orcid.org/0000-0001-7259-9751S. C. Chan3https://orcid.org/0000-0001-7212-4182Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong KongDepartment of Electrical and Electronic Engineering, The University of Hong Kong, Hong KongDepartment of Electrical and Electronic Engineering, The University of Hong Kong, Hong KongDepartment of Electrical and Electronic Engineering, The University of Hong Kong, Hong KongPerson re-identification problems usually suffer from large subject appearance variations and limited training data. This paper proposes a novel physically motivated Color/Illuminance-Aware data-augmentation (CIADA) scheme and a style-adaptive fusion approach to address these issues. The CIADA scheme estimates the color/illuminance distribution from the training data via manifold learning and generates new samples under different color/illuminance perturbations to better capture objects’ appearance for mitigating the small-sample-size and color variation problems. A Color/Illuminance Aware Feature Augmentation (CIAFA) approach, which is applicable to state-of-the-art features and metric learning algorithms, is then proposed to integrate the features generated by the augmented samples for metric learning. A new Color/Illuminance-Aware Style Fusion (CIASF) scheme, which allows the learning and matching process to be performed independently on each pair of datasets generated for estimating a set of ‘local’ distance functions, is also proposed. A canonical correlation analysis-based weighting scheme is developed to fuse these local distances to an overall distance for recognition. This reduces the memory requirement and complexity over the original CIAFA. Experiments on common datasets show that the proposed methodologies substantially improve the performance of state-of-the-art subspace learning algorithms. It is applicable to both small and large datasets with hand-craft and deep features.https://ieeexplore.ieee.org/document/9497884/Data augmentationlocal metric learningsmall sample size problemperson re-identification |
spellingShingle | Zhouchi Lin Chenyang Liu Wenbo Qi S. C. Chan A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification IEEE Access Data augmentation local metric learning small sample size problem person re-identification |
title | A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification |
title_full | A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification |
title_fullStr | A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification |
title_full_unstemmed | A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification |
title_short | A Color/Illuminance Aware Data Augmentation and Style Adaptation Approach to Person Re-Identification |
title_sort | color illuminance aware data augmentation and style adaptation approach to person re identification |
topic | Data augmentation local metric learning small sample size problem person re-identification |
url | https://ieeexplore.ieee.org/document/9497884/ |
work_keys_str_mv | AT zhouchilin acolorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT chenyangliu acolorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT wenboqi acolorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT scchan acolorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT zhouchilin colorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT chenyangliu colorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT wenboqi colorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification AT scchan colorilluminanceawaredataaugmentationandstyleadaptationapproachtopersonreidentification |