基于多尺度特征的行人重识别属性提取新方法

doi:10.13229/j.cnki.jdxbgxb.20220327

Abstract

Abstract:

At present， person re-identification （ReID） still exists some issues to solve， such as background confusion， blocking interference， pose misalignment， and insufficient explanation. In this paper， a new method based on multi-scale feature extraction is proposed for person ReID. Specifically， using the feature learning module to learn attributes and convolving by attribute grouping， which could enhance the interpretability of the model and reduce the impact of interference features by comparing the intermediate property characteristics. Compared with other methods， the new method was tested on three commonly used public datasets. The detailed experiments demonstrate that the new method can efficiently extract semantic information from the middle layer and outperforms other methods tested across datasets.

Key words: computer application, interpretability, person re-identification, multi-scale features

CLC Number:

TP391.4

Peng YU,Yan PIAO. New method for extracting person re-identification attributes based on multi-scale features[J].Journal of Jilin University(Engineering and Technology Edition), 2023, 53(4): 1155-1162.

Figures/Tables 8

Fig.1

Table 1

Fig.2

Table 2

Fig.3

Table 3

Table 4

Table 5

References 27

1	Alex K, Ilya S, Geoffrey E H. Imagenet classification with deep convolutional neural networks[J]. Advances in neural information processing systems, 2017, 60(6): 84-90.
2	Zhang Zhi-zheng, Lan Cui-ling, Zeng Wen-jun, et al. Relation-aware global attention for person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2020: 3186-3195.
3	Chen Guang-yi, Lin Chun-ze, Ren Liang-liang, et al. Self-critical attention learning for person re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 2019: 9637-9646.
4	Fang Peng-fei, Zhou Jie-ming, Kumar Roy Soumava, et al. Bilinear attention networks for person retrieval[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 2019: 8030-8039.
5	Sun Yi-fan, Zheng Liang, Yang Yi, et al. Beyond part models: person retrieval with refined part pooling (and a strong convolutional baseline)[C]//Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany, 2018: 501-518.
6	Zhou Kai-yang, Yang Yong-xin, Cavallaro Andrea, et al. Omni-scale feature learning for person re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 2019: 3702-3712.
7	Chen Xiao-dong, Liu Xin-chen, Liu Wu, et al. Explainable person re-identification with attribute-guided metric distillation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal, Canada, 2021: 11813-11822.
8	Divyansh G, Netra P, Thomas T, et al. Towards explainable person re-identification[C]//2021 IEEE Symposium Series on Computational Intelligence (SSCI), Orlando, USA, 2021: 1-8.
9	Chen Guang-yi, Gu Tian-pei, Lu Ji-wen, et al. Person re-identification via attention pyramid[J]. IEEE Transactions on Image Processing, 2021, 30: 7663-7676.
10	Wang Guan-shuo, Yuan Yu-feng, Chen Xiong, et al. Learning discriminative features with multiple granularities for person re-identification[C]//Proceedings of the 26th ACM International Conference on Multimedia, Seoul, Korea(South), 2018: 274-282.
11	Feng Zheng, Cheng Deng, Xing Sun, et al. Pyramidal person re-identification via multi-loss dynamic training[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 8514-8522.
12	本杰明•里贝特. 心智时间: 意识中的时间因素[M]: 李恒熙,李恒威,罗慧怡译.杭州: 浙江大学出版社, 2013:55-56.
13	Han Yi-zeng, Huang Gao, Song Shi-ji, et al. Dynamic neural networks: a survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 44(11): 7436-7456.
14	Teerapittayanon S, McDanel B, Kung H T. Branchynet: fast inference via early exiting from deep neural networks[C]//The 23rd International Conference on Pattern Recognition (ICPR), Cancun, Mexico, 2016: 2464-2469.
15	Hermans Alexander, Beyer Lucas, Leibe Bastian. In defense of the triplet loss for person re-identification. arXiv 2017[J/OL].[2017-04-01]. .
16	Zheng Liang, Shen Li-yue, Tian Lu, et al. Scalable person re-identification: a benchmark[C]//Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile, 2015: 1116-1124.
17	Zhong Zhun, Zheng Liang, Cao Dong-lin, et al. Re-ranking person re-identification with k-reciprocal encoding[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 1318-1327.
18	Ergys R, Francesco S, Roger Z, et al. Performance measures and a data set for multi-target, multi-camera tracking[C]//European Conference on Computer Vision, Amsterdam, The Netherlands, 2016: 17-35.
19	Lin Yu-tian, Zheng Liang, Zheng Zhe-dong, et al. Improving person re-identification by attribute and identity learning[J]. Pattern Recognition, 2019, 95: 151-161.
20	Jia Deng, Wei Dong, Socher Richard, et al. Imagenet: a large-scale hierarchical image database[C]//IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, 2009: 248-255.
21	Hyunjong P, Bumsub H. Relation network for person re-identification[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(7): 11839-11847.
22	Sun Yi-fan, Xu Qin, Li Ya-li, et al. Perceive where to focus: learning visibility-aware part-level features for partial person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 393-402.
23	Tay C-P, Roy S, Yap K-H. Aanet: attribute attention network for person re-identifications[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 7134-7143.
24	Yang Wen-jie, Huang Hou-jing, Zhang Zhang, et al. Towards rich feature discovery with class activation maps augmentation for person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 1389-1398.
25	Chen Bing-hui, Deng Wei-hong, Hu Jia-ni. Mixed high-order attention network for person re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 2019: 371-381.
26	Chen Xue-song, Fu Can-miao, Zhao Yong, et al. Salience-guided cascaded suppression network for person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2020: 3300-3310.
27	Zhao Shi-zhen, Gao Chang-xin, Zhang Jun, et al. Do not disturb me: person re-identification under the interference of other pedestrians[C]//European Conference on Computer Vision, Glasgow, UK, 2020: 647-663.

Related Articles 15

[1]	Yu JIANG,Jia-zheng PAN,He-huai CHEN,Ling-zhi FU,Hong QI. Segmentation-based detector for traditional Chinese newspaper [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(4): 1146-1154.
[2]	Zhen-yu WU,Xiao-fei LIU,Yi-pu WANG. Trajectory planning of unmanned system based on DKRRT^*⁃APF algorithm [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(3): 781-791.
[3]	Bo TAO,Fu-wu YAN,Zhi-shuai YIN,Dong-mei WU. 3D object detection based on high⁃precision map enhancement [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(3): 802-809.
[4]	Shan XUE,Ya-liang ZHANG,Qiong-ying LYU,Guo-hua CAO. Anti⁃unmanned aerial vehicle system object detection algorithm under complex background [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(3): 891-901.
[5]	Hong-yang PAN,Zhao LIU,Bo YANG,Geng SUN,Yan-heng LIU. Overview of swarm intelligence methods for unmanned aerial vehicle systems based on new⁃generation information technology [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(3): 629-642.
[6]	Ying HE,Jun-song FAN,Wei WANG,Geng SUN,Yan-heng LIU. Joint optimization of secure communication and trajectory planning in unmanned aerial vehicle air⁃to⁃ground [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(3): 913-922.
[7]	Peng GUO,Wen-chao ZHAO,Kun LEI. Dual⁃resource constrained flexible job shop optimal scheduling based on an improved Jaya algorithm [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(2): 480-487.
[8]	Jin-Zhen Liu,Guo-Hui Gao,Hui Xiong. Multi⁃scale attention network for brain tissue segmentation [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(2): 576-583.
[9]	Xian-yu QI,Wei WANG,Lin WANG,Yu-fei ZHAO,Yan-peng DONG. Semantic topological map building with object semantic grid map [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(2): 569-575.
[10]	Xiao-hu SHI,Jia-qi WU,Chun-guo WU,Shi CHENG,Xiao-hui WENG,Zhi-yong CHANG. Residual network based curve enhanced lane detection method [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(2): 584-592.
[11]	Fu-heng QU,Tian-yu DING,Yang LU,Yong YANG,Ya-ting HU. Fast image codeword search algorithm based on neighborhood similarity [J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(8): 1865-1871.
[12]	Tian BAI,Ming-wei XU,Si-ming LIU,Ji-an ZHANG,Zhe WANG. Dispute focus identification of pleading text based on deep neural network [J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(8): 1872-1880.
[13]	Feng-feng ZHOU,Hai-yang ZHU. SEE： sense EEG⁃based emotion algorithm via three⁃step feature selection strategy [J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(8): 1834-1841.
[14]	Gui-he QIN,Jun-feng HUANG,Ming-hui SUN. Text input based on two⁃handed keyboard in virtual environment [J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(8): 1881-1888.
[15]	Feng-feng ZHOU,Yi-chi ZHANG. Unsupervised feature engineering algorithm BioSAE based on sparse autoencoder [J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(7): 1645-1656.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Comments

Recommended 0

No Suggested Reading articles found!

阶段		输出尺寸	OSNet	阶段	输出尺寸	MLLANet
	conv1	128×64， 64	7×7 conv， stride 2
	conv1	64×32， 64	3×3 max pool， stride 2
1层	conv2	64×32， 256	bottleneck × 2
	transition	64×32， 256	1×1 conv
	transition	32×16， 256	2×2 average pool， stride 2
	MLL0	32×16， 256	1×1 conv	MLLA0	32×16， 256×30+30	1×1 conv
		1×1， 256	global average pool		1×1， 256×30+30	global average pool
		1×1， 256	fc		1×1， 256×30+30	fc
2层	conv3	32×16， 384	bottleneck × 2
	transition	32×16， 384	1×1 conv
	transition	16×8， 384	2×2 average pool， stride 2
	MLL1	16×8， 384	1×1 conv	MLLA1	16×8， 384 ×30+30	1×1 conv
		1×1， 384	global average pool		1×1， 384 ×30+30	global average pool
		1×1， 384	fc		1×1， 384 ×30+30	fc
3层	conv4	16×8， 512	bottleneck × 2
	MLL2	16×8， 512	1×1 conv	MLLA2	16×8， 512 ×30+30	1×1 conv
		1×1， 512	global average pool		1×1， 512 ×30+30	global average pool
		1×1， 512	fc		1×1， 512 ×30+30	fc

	2层	3层	4层
飞机	313	415	272
汽车	340	403	257
鸟	338	418	244
猫	303	457	240
鹿	329	436	235
狗	307	460	233
蛙	325	419	256
马	338	415	247
船	295	450	255
卡车	312	427	261

模型	权重	平均精度	前1次命中的准确率	前5次命中的准确率
OSNet	预训练	84.9	94.8	96.8
OSNet	无预训练	63.0	83.5	93.3
OSNet+MLL	预训练	86.7	95.9	97.4
OSNet+MLL	无预训练	72.5	89.3	95.5
GCP	预训练	88.9	95.2	98.2
GCP	无预训练	85.3	93.9	97.9
GCP+MLL	预训练	89.7	95.4	98.5
GCP+MLL	无预训练	87.7	94.4	98.1

方法	主干网	Market?1501		DukeMTMC?ReID			CUHK03
方法	主干网	mAP	Rank?1	mAP	Rank?1	mAP	Rank?1
PCB+RPP^［5］	ResNet?50	81.6	93.8	69.2	83.3	57.5	63.7
MGN^［10］	ResNet?50	86.9	95.6	78.4	88.7	66.0	66.8
VMP^［22］	ResNet?50	80.8	93.0	72.6	83.6	-	-
OSNet^［6］	OSNet	84.9	94.8	73.5	88.6	67.8	72.3
AAnet^［23］	ResNet?50	82.5	93.9	72.6	86.4	-	-
CAMA^［24］	ResNet?50	84.5	94.7	-	-	64.2	66.6
BAT?net^［4］	ResNet?50	85.5	94.1	77.3	87.7	73.2	76.2
MHN^［25］	ResNet?50	85.0	95.1	77.2	89.1	65.4	71.7
SCSN^［26］	ResNet?50	88.5	95.7	79.0	90.1	80.2	84.1
GCP^［21］	ResNet?50	88.9	95.2	78.6	89.7	75.6	77.9
RGA?SC^［2］	ResNet?50	88.4	96.1	-	-	74.5	79.6
?RGA?S^［2］	ResNet?50	88.0	95.7	77.5	88.8	72.7	78.1
PISNet^［27］	ResNet?50	87.1	95.6	78.7	88.8	-	-
APNet?S^［9］	ResNet?50	89.0	96.1	78.8	89.3	78.1	80.9
APNet?C^［9］	ResNet?50	90.5	96.2	81.5	90.4	81.5	83.0
MLLNet	OSNet	86.7	95.9	74.6	89.2	69.7	73.4
MLLANet	OSNet	87.8	96.1	75.1	89.5	-	-
MLLNet2	APNet?C	89.9	95.9	81.7	90.5	81.7	83.1
MLLNet3	GCP	89.7	95.4	79.5	90.0	76.9	78.7

训练集	测试集	模型	平均精度	Rank?1	Rank?5	Rank?10
Market	Duke	APNet?c3	22.8	37.7	52.4	59.0
Market	Duke	OSNet	26.7	48.5	62.3	67.4
Market	Duke	OSNet_MLL	26.9	49.5	63.1	68.0
Market	Duke	OSNet_MLLA	27.1	49.6	63.7	69.0
Duke	Market	APNet?c3	23.7	50.9	66.6	72.6
Duke	Market	OSNet	26.1	57.7	73.7	80.0
Duke	Market	OSNet_MLL	26.2	58.0	73.4	82.5
Duke	Market	OSNet_MLLA	26.8	59.8	75.6	81.5

New method for extracting person re-identification attributes based on multi-scale features

RICH HTML

PDF (PC)