A biologically inspired separable learning vision model for real-time traffic object perception in Dark

Fast and accurate object perception in low-light traffic scenes has attracted increasing attention. However, due to severe illumination degradation and the lack of reliable visual cues, existing perception models and methods struggle to quickly adapt to and accurately predict in low-light environmen...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Expert systems with applications Ročník 297; s. 129529
Hlavní autori: Li, Hulin, Ren, Qiliang, Li, Jun, Wei, Hanbing, Liu, Zheng, Fan, Linfang
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Ltd 01.02.2026
Predmet:
ISSN:0957-4174
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Fast and accurate object perception in low-light traffic scenes has attracted increasing attention. However, due to severe illumination degradation and the lack of reliable visual cues, existing perception models and methods struggle to quickly adapt to and accurately predict in low-light environments. Moreover, there is the absence of available large-scale benchmark specifically focused on low-light traffic scenes. To bridge this gap, we introduce a physically grounded illumination degradation method tailored to real-world low-light settings and construct Dark-traffic, the largest densely annotated dataset to date for low-light traffic scenes, supporting object detection, instance segmentation, and optical flow estimation. We further propose the Separable Learning Vision Model (SLVM), a biologically inspired framework designed to enhance perception under adverse lighting. SLVM integrates four key components: a light-adaptive pupillary mechanism for illumination-sensitive feature extraction, a feature-level separable learning strategy for efficient representation, task-specific decoupled branches for multi-task separable learning, and a spatial misalignment-aware fusion module for precise multi-feature alignment. Extensive experiments demonstrate that SLVM achieves state-of-the-art performance with reduced computational overhead. Notably, it outperforms RT-DETR by 11.2 percentage points in detection, YOLOv12 by 6.1 percentage points in instance segmentation, and reduces endpoint error (EPE) of baseline by 12.37 % on Dark-traffic. On the LIS benchmark, the end-to-end trained SLVM surpasses Swin Transformer + EnlightenGAN and ConvNeXt-T + EnlightenGAN by an average of 11 percentage points across key metrics, and exceeds Mask RCNN (with light enhancement) by 3.1 percentage points. The Dark-traffic dataset and complete code is released at https://github.com/alanli1997/slvm.
AbstractList Fast and accurate object perception in low-light traffic scenes has attracted increasing attention. However, due to severe illumination degradation and the lack of reliable visual cues, existing perception models and methods struggle to quickly adapt to and accurately predict in low-light environments. Moreover, there is the absence of available large-scale benchmark specifically focused on low-light traffic scenes. To bridge this gap, we introduce a physically grounded illumination degradation method tailored to real-world low-light settings and construct Dark-traffic, the largest densely annotated dataset to date for low-light traffic scenes, supporting object detection, instance segmentation, and optical flow estimation. We further propose the Separable Learning Vision Model (SLVM), a biologically inspired framework designed to enhance perception under adverse lighting. SLVM integrates four key components: a light-adaptive pupillary mechanism for illumination-sensitive feature extraction, a feature-level separable learning strategy for efficient representation, task-specific decoupled branches for multi-task separable learning, and a spatial misalignment-aware fusion module for precise multi-feature alignment. Extensive experiments demonstrate that SLVM achieves state-of-the-art performance with reduced computational overhead. Notably, it outperforms RT-DETR by 11.2 percentage points in detection, YOLOv12 by 6.1 percentage points in instance segmentation, and reduces endpoint error (EPE) of baseline by 12.37 % on Dark-traffic. On the LIS benchmark, the end-to-end trained SLVM surpasses Swin Transformer + EnlightenGAN and ConvNeXt-T + EnlightenGAN by an average of 11 percentage points across key metrics, and exceeds Mask RCNN (with light enhancement) by 3.1 percentage points. The Dark-traffic dataset and complete code is released at https://github.com/alanli1997/slvm.
ArticleNumber 129529
Author Li, Jun
Fan, Linfang
Wei, Hanbing
Li, Hulin
Liu, Zheng
Ren, Qiliang
Author_xml – sequence: 1
  givenname: Hulin
  surname: Li
  fullname: Li, Hulin
  email: alan@mails.cqjtu.edu.cn
  organization: School of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
– sequence: 2
  givenname: Qiliang
  surname: Ren
  fullname: Ren, Qiliang
  email: qlren@cqjtu.edu.cn
  organization: School of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
– sequence: 3
  givenname: Jun
  surname: Li
  fullname: Li, Jun
  email: cqleejun@cqjtu.edu.cn
  organization: School of Mechatronics and Vehicle Engineering, Chongqing Jiaotong University, Chongqing 400074, China
– sequence: 4
  givenname: Hanbing
  surname: Wei
  fullname: Wei, Hanbing
  email: hbwei@cqjtu.edu.cn
  organization: School of Mechatronics and Vehicle Engineering, Chongqing Jiaotong University, Chongqing 400074, China
– sequence: 5
  givenname: Zheng
  surname: Liu
  fullname: Liu, Zheng
  email: zheng.liu@ubc.ca
  organization: School of Engineering, University of British Columbia, Okanagan, Kelowna BC V1V 1V7, Canada
– sequence: 6
  givenname: Linfang
  surname: Fan
  fullname: Fan, Linfang
  email: linfangfan@cqjtu.edu.cn
  organization: School of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
BookMark eNp90L1OwzAQwHEPRaItvACTXyDBduq4kViq8lGkSiwwWxfnXDm4dmRHRX17GpWZ6Zb7n06_BZmFGJCQB85Kznj92JeYf6AUTMiSi0aKZkbmrJGqWHG1uiWLnHvGuGJMzclhQ1sXfTw4A96fqQt5cAk7mnGABK1H6hFScOFATy67GOgxduipjYkmBF-M7oh0TGCtMzS2PZqRDpgMDuO07QJ9hvR9R24s-Iz3f3NJvl5fPre7Yv_x9r7d7AsjZDUWdd1WDRN1hdVacY6A0oKyjeCqXWOl7Ko1XWOaGiwa0a1BSsUMg06KCoWAaknE9a5JMeeEVg_JHSGdNWd64tG9nnj0xKOvPJfo6Rrh5bOTw6SzcRgMdhcKM-ouuv_yX_CLdEw
Cites_doi 10.1109/CVPR.2018.00347
10.1109/CVPRW.2017.66
10.1126/science.156.3775.636
10.1109/CVPR.2016.91
10.1109/CVPR42600.2020.00185
10.1109/CVPR42600.2020.00678
10.1109/CVPR.2019.01099
10.1109/CVPR52688.2022.00795
10.1109/CVPR52733.2024.01605
10.1109/CVPR.2017.106
10.1109/TIP.2021.3051462
10.1145/3343031.3350926
10.1016/j.isprsjprs.2021.11.010
10.1109/CVPR.2016.90
10.1007/s11554-024-01436-6
10.1109/CVPR52688.2022.01167
10.1038/s41467-024-50488-6
10.1109/CVPR42600.2020.00982
10.1109/CVPR.2016.308
10.1038/s41593-019-0520-2
10.1016/j.compbiomed.2024.107917
10.1109/ICCV.2017.322
10.1109/CVPR52688.2022.00135
10.1038/s41586-024-07566-y
10.1007/978-3-030-01264-9_8
10.1109/CVPRW50498.2020.00203
10.1007/s11263-023-01808-8
10.1109/ICCV48922.2021.00986
10.1109/ICCV.2019.00925
10.1038/s41593-019-0470-8
10.1109/NCC.2012.6176791
10.1016/0166-2236(83)90190-X
10.1109/CVPR52733.2024.01440
10.1007/978-3-319-10602-1_48
10.1007/978-3-031-72751-1_1
10.1016/j.cviu.2018.10.010
10.1109/CVPR42600.2020.01020
10.1109/CVPR46437.2021.00349
ContentType Journal Article
Copyright 2025 Elsevier Ltd
Copyright_xml – notice: 2025 Elsevier Ltd
DBID AAYXX
CITATION
DOI 10.1016/j.eswa.2025.129529
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
ExternalDocumentID 10_1016_j_eswa_2025_129529
S0957417425031446
GroupedDBID --K
--M
.DC
.~1
0R~
13V
1B1
1RT
1~.
1~5
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
9JO
AAAKF
AABNK
AAEDT
AAEDW
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AARIN
AATTM
AAXKI
AAXUO
AAYFN
AAYWO
ABBOA
ABFNM
ABJNI
ABMAC
ABMVD
ABUCO
ABUFD
ACDAQ
ACGFS
ACHRH
ACLOT
ACNTT
ACRLP
ACVFH
ACZNC
ADBBV
ADCNI
ADEZE
ADTZH
AEBSH
AECPX
AEIPS
AEKER
AENEX
AEUPX
AFJKZ
AFPUW
AFTJW
AGHFR
AGUBO
AGUMN
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIGII
AIIUN
AIKHN
AITUG
AKBMS
AKRWK
AKYEP
ALEQD
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
ANKPU
AOUOD
APLSM
APXCP
AXJTR
BJAXD
BKOJK
BLXMC
BNSAS
CS3
DU5
EBS
EFJIC
EFKBS
EFLBG
EO8
EO9
EP2
EP3
F5P
FDB
FIRID
FNPLU
FYGXN
G-Q
GBLVA
GBOLZ
HAMUX
IHE
J1W
JJJVA
KOM
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
ROL
RPZ
SDF
SDG
SDP
SDS
SES
SEW
SPC
SPCBC
SSB
SSD
SSL
SST
SSV
SSZ
T5K
TN5
~G-
~HD
29G
9DU
AAAKG
AAQXK
AAYXX
ABKBG
ABWVN
ABXDB
ACNNM
ACRPL
ADJOM
ADMUD
ADNMO
AGQPQ
ASPBG
AVWKF
AZFZN
CITATION
EJD
FEDTE
FGOYB
G-2
HLZ
HVGLF
HZ~
LG9
LY1
LY7
M41
R2-
SBC
SET
WUQ
XPP
ZMT
ID FETCH-LOGICAL-c253t-66b390263e38711eae5fa7f9217b8e37f4bcd9c96afec2d8a5570c0ad523e22a3
ISICitedReferencesCount 1
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001568955900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0957-4174
IngestDate Sat Nov 29 06:52:01 EST 2025
Sat Nov 29 17:05:08 EST 2025
IsPeerReviewed true
IsScholarly true
Keywords Deep learning
Instance segmentation
Bio-inspired vision
Low-light traffic
Object detection
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c253t-66b390263e38711eae5fa7f9217b8e37f4bcd9c96afec2d8a5570c0ad523e22a3
ParticipantIDs crossref_primary_10_1016_j_eswa_2025_129529
elsevier_sciencedirect_doi_10_1016_j_eswa_2025_129529
PublicationCentury 2000
PublicationDate 2026-02-01
2026-02-00
PublicationDateYYYYMMDD 2026-02-01
PublicationDate_xml – month: 02
  year: 2026
  text: 2026-02-01
  day: 01
PublicationDecade 2020
PublicationTitle Expert systems with applications
PublicationYear 2026
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Khanam R, Hussain M. Yolov11: An overview of the key architectural enhancements. arXiv preprint arXiv:2410.17725, 2024.
Cheng B, Misra I, Schwing A G, et al. Masked-attention mask transformer for universal image segmentation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 1290–1299.
Lin T Y, Maire M, Belongie S, et al. Microsoft coco: Common objects in context[C]//Computer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part v 13. Springer International Publishing, 2014: 740–755.
Jocher G, Chaurasia A, Qiu J (2023). YOLO by Ultralytics (Version 8.0.0). Computer software. GitHub. Retrieved from https://github.com/ultralytics/ultralytics.
Mandelbrot (b0160) 1967; 156
Shumailov, Shumaylov, Zhao (b0200) 2024; 631
Li J, Li B, Tu Z, et al. Light the night: A multi-condition diffusion framework for unpaired low-light enhancement in autonomous driving. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2024: 15205-15215.
Li, Li, Wang (b0125) 2024; 15
Liu Z, Mao H, Wu C Y, et al. A convnet for the 2020s. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 11976-11986.
De Brabandere B, Neven D, Van Gool L. Semantic instance segmentation for autonomous driving. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. 2017: 7-9.
Chen, Wang, Guo (b0025) 2023; 36
Li, Li, Wei (b0120) 2024; 21
Lin T Y, Dollár P, Girshick R, et al. Feature pyramid networks for object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2117–2125.
Loh, Chan (b0150) 2019; 178
Pongrácz (b0180) 2014
Parthasarathy S, Sankaran P. An automated multi scale retinex with color restoration for image enhancement. In: 2012 National Conference on Communications (NCC). IEEE, 2012: 1–5.
Rahman, Jobson, Woodell (b0185) 1996
Ge Z, Liu S, Wang F, et al. Yolox: Exceeding yolo series in 2021. arXiv preprint arXiv:2107.08430, 2021.
Kirillov A, Wu Y, He K, et al. Pointrend: Image segmentation as rendering. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 9799-9808.
Tian Y, Ye Q, Doermann D. Yolov12: Attention-centric real-time object detectors. arXiv preprint arXiv:2502.12524, 2025.
Geiger, Lenz, Urtasun (b0060) 2012
Liu Z, Lin Y, Cao Y, et al. Swin transformer: Hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF international conference on computer vision. 2021: 10012–10022.
Wang, Chen, Liu (b0220) 2024; 37
Xu H, Zhang J, Cai J, et al. Gmflow: Learning optical flow via global matching. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 8121–8130.
Richards, Lillicrap, Beaudoin (b0195) 2019; 22
Wang C Y, Liao H Y M, Wu Y H, et al. CSPNet: A new backbone that can enhance learning capability of CNN. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. 2020: 390-391.
Chen, Fu, Wei (b0020) 2023; 131
Zhang Z, Gupta A, Jiang H, et al. Neuflow v2: High-efficiency optical flow estimation on edge devices. arXiv preprint arXiv:2408.10161, 2024.
Zhang, Nex, Kerle (b0260) 2022; 183
Chen C, Chen Q, Xu J, et al. Learning to see in the dark. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 3291-3300.
Guo C, Li C, Guo J, et al. Zero-reference deep curve estimation for low-light image enhancement. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 1780-1789.
Lamba M, Mitra K. Restoring extremely dark images in real time. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021: 3487-3497.
Chen, Zhang, Chen (b0030) 2024; 170
Ma N, Zhang X, Zheng H T, et al. Shufflenet v2: Practical guidelines for efficient cnn architecture design. In: Proceedings of the European conference on computer vision (ECCV). 2018: 116-131.
He K, Zhang X, Ren S, et al. Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770–778.
Wu Y, Chen Y, Yuan L, et al. Rethinking classification and localization for object detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 10186–10195.
Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 779–788.
Jiang, Gong, Liu (b0085) 2021; 30
Choi, Han, Wang (b0040) 2023; 36
He K, Gkioxari G, Dollár P, et al. Mask r-cnn. In: Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.
Wu B, Dai X, Zhang P, et al. Fbnet: Hardware-aware efficient convnet design via differentiable neural architecture search. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 10734-10742.
Niv (b0170) 2019; 22
Burch, Bailey (b0010) 2008
Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020.
Zheng Y, Zhang M, Lu F. (2020). Optical flow in the dark. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition: 6749-6757.
Li (b0110) 2024
Wang C Y, Yeh I H, Mark Liao H Y. Yolov9: Learning what you want to learn using programmable gradient information[C]//European conference on computer vision. Cham: Springer Nature Switzerland, 2024: 1–21.
Bolya D, Zhou C, Xiao F, et al. Yolact: Real-time instance segmentation. In: Proceedings of the IEEE/CVF international conference on computer vision. 2019: 9157–9166.
Zhang Y, Zhang J, Guo X. Kindling the darkness: A practical low-light image enhancer. In: Proceedings of the 27th ACM international conference on multimedia. 2019: 1632–1640.
Mishkin, Ungerleider, Macko (b0165) 1983; 6
Simonyan K, Zisserman A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
Zhao Y, Lv W, Xu S, et al. Detrs beat yolos on real-time object detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2024: 16965-16974.
Szegedy C, Vanhoucke V, Ioffe S, et al. Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 2818-2826.
10.1016/j.eswa.2025.129529_b0215
Wang (10.1016/j.eswa.2025.129529_b0220) 2024; 37
10.1016/j.eswa.2025.129529_b0255
10.1016/j.eswa.2025.129529_b0135
Shumailov (10.1016/j.eswa.2025.129529_b0200) 2024; 631
10.1016/j.eswa.2025.129529_b0015
Chen (10.1016/j.eswa.2025.129529_b0025) 2023; 36
Jiang (10.1016/j.eswa.2025.129529_b0085) 2021; 30
Li (10.1016/j.eswa.2025.129529_b0110) 2024
10.1016/j.eswa.2025.129529_b0065
Chen (10.1016/j.eswa.2025.129529_b0030) 2024; 170
10.1016/j.eswa.2025.129529_b0100
10.1016/j.eswa.2025.129529_b0265
Li (10.1016/j.eswa.2025.129529_b0120) 2024; 21
Loh (10.1016/j.eswa.2025.129529_b0150) 2019; 178
10.1016/j.eswa.2025.129529_b0140
10.1016/j.eswa.2025.129529_b0105
10.1016/j.eswa.2025.129529_b0145
Pongrácz (10.1016/j.eswa.2025.129529_b0180) 2014
10.1016/j.eswa.2025.129529_b0225
Geiger (10.1016/j.eswa.2025.129529_b0060) 2012
Choi (10.1016/j.eswa.2025.129529_b0040) 2023; 36
Li (10.1016/j.eswa.2025.129529_b0125) 2024; 15
10.1016/j.eswa.2025.129529_b0190
10.1016/j.eswa.2025.129529_b0070
10.1016/j.eswa.2025.129529_b0075
10.1016/j.eswa.2025.129529_b0230
10.1016/j.eswa.2025.129529_b0155
10.1016/j.eswa.2025.129529_b0270
Richards (10.1016/j.eswa.2025.129529_b0195) 2019; 22
Zhang (10.1016/j.eswa.2025.129529_b0260) 2022; 183
10.1016/j.eswa.2025.129529_b0035
10.1016/j.eswa.2025.129529_b0235
10.1016/j.eswa.2025.129529_b0115
Niv (10.1016/j.eswa.2025.129529_b0170) 2019; 22
Chen (10.1016/j.eswa.2025.129529_b0020) 2023; 131
10.1016/j.eswa.2025.129529_b0240
10.1016/j.eswa.2025.129529_b0045
Burch (10.1016/j.eswa.2025.129529_b0010) 2008
10.1016/j.eswa.2025.129529_b0205
Mandelbrot (10.1016/j.eswa.2025.129529_b0160) 1967; 156
Rahman (10.1016/j.eswa.2025.129529_b0185) 1996
10.1016/j.eswa.2025.129529_b0245
10.1016/j.eswa.2025.129529_b0005
10.1016/j.eswa.2025.129529_b0090
Mishkin (10.1016/j.eswa.2025.129529_b0165) 1983; 6
10.1016/j.eswa.2025.129529_b0130
10.1016/j.eswa.2025.129529_b0175
10.1016/j.eswa.2025.129529_b0055
10.1016/j.eswa.2025.129529_b0210
10.1016/j.eswa.2025.129529_b0050
10.1016/j.eswa.2025.129529_b0095
10.1016/j.eswa.2025.129529_b0250
References_xml – reference: Li J, Li B, Tu Z, et al. Light the night: A multi-condition diffusion framework for unpaired low-light enhancement in autonomous driving. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2024: 15205-15215.
– volume: 36
  start-page: 7050
  year: 2023
  end-page: 7064
  ident: b0025
  article-title: Vanillanet: The power of minimalism in deep learning
  publication-title: Advances in Neural Information Processing Systems
– reference: Lin T Y, Dollár P, Girshick R, et al. Feature pyramid networks for object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2117–2125.
– reference: Liu Z, Mao H, Wu C Y, et al. A convnet for the 2020s. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 11976-11986.
– volume: 631
  start-page: 755
  year: 2024
  end-page: 759
  ident: b0200
  article-title: AI models collapse when trained on recursively generated data
  publication-title: Nature
– reference: Simonyan K, Zisserman A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
– reference: Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 779–788.
– reference: Parthasarathy S, Sankaran P. An automated multi scale retinex with color restoration for image enhancement. In: 2012 National Conference on Communications (NCC). IEEE, 2012: 1–5.
– volume: 6
  start-page: 414
  year: 1983
  end-page: 417
  ident: b0165
  article-title: Object vision and spatial vision: Two cortical pathways
  publication-title: Trends in neurosciences
– reference: Tian Y, Ye Q, Doermann D. Yolov12: Attention-centric real-time object detectors. arXiv preprint arXiv:2502.12524, 2025.
– reference: Wang C Y, Liao H Y M, Wu Y H, et al. CSPNet: A new backbone that can enhance learning capability of CNN. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. 2020: 390-391.
– year: 2008
  ident: b0010
  article-title: How Dogs Learn
– volume: 22
  start-page: 1544
  year: 2019
  end-page: 1553
  ident: b0170
  article-title: Learning task-state representations
  publication-title: Nature Neuroscience
– reference: De Brabandere B, Neven D, Van Gool L. Semantic instance segmentation for autonomous driving. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. 2017: 7-9.
– reference: Liu Z, Lin Y, Cao Y, et al. Swin transformer: Hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF international conference on computer vision. 2021: 10012–10022.
– reference: Chen C, Chen Q, Xu J, et al. Learning to see in the dark. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 3291-3300.
– reference: Jocher G, Chaurasia A, Qiu J (2023). YOLO by Ultralytics (Version 8.0.0). Computer software. GitHub. Retrieved from https://github.com/ultralytics/ultralytics.
– volume: 178
  start-page: 30
  year: 2019
  end-page: 42
  ident: b0150
  article-title: Getting to know low-light images with the exclusively dark dataset
  publication-title: Computer Vision and Image Understanding
– start-page: 3354
  year: 2012
  end-page: 3361
  ident: b0060
  article-title: Are we ready for autonomous driving? the kitti vision benchmark suite
  publication-title: 2012 IEEE conference on computer vision and pattern recognition
– volume: 170
  year: 2024
  ident: b0030
  article-title: Accurate leukocyte detection based on deformable-DETR and multi-level feature fusion for aiding diagnosis of blood diseases
  publication-title: Computers in Biology and Medicine
– volume: 131
  start-page: 2198
  year: 2023
  end-page: 2218
  ident: b0020
  article-title: Instance segmentation in the dark
  publication-title: International Journal of Computer Vision
– volume: 36
  start-page: 50408
  year: 2023
  end-page: 50428
  ident: b0040
  article-title: A dual-stream neural network explains the functional segregation of dorsal and ventral visual pathways in human brains
  publication-title: Advances in Neural Information Processing Systems
– reference: He K, Gkioxari G, Dollár P, et al. Mask r-cnn. In: Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.
– reference: Zhang Z, Gupta A, Jiang H, et al. Neuflow v2: High-efficiency optical flow estimation on edge devices. arXiv preprint arXiv:2408.10161, 2024.
– volume: 21
  start-page: 62
  year: 2024
  ident: b0120
  article-title: Slim-neck by GSConv: A lightweight-design for real-time detector architectures
  publication-title: Journal of Real-Time Image Processing
– volume: 22
  start-page: 1761
  year: 2019
  end-page: 1770
  ident: b0195
  article-title: A deep learning framework for neuroscience
  publication-title: Nature Neuroscience
– reference: Zhao Y, Lv W, Xu S, et al. Detrs beat yolos on real-time object detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2024: 16965-16974.
– reference: Wu B, Dai X, Zhang P, et al. Fbnet: Hardware-aware efficient convnet design via differentiable neural architecture search. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 10734-10742.
– reference: Lamba M, Mitra K. Restoring extremely dark images in real time. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021: 3487-3497.
– start-page: 249
  year: 2014
  end-page: 293
  ident: b0180
  article-title: Social learning in dogs
  publication-title: The Social Dog
– reference: Bolya D, Zhou C, Xiao F, et al. Yolact: Real-time instance segmentation. In: Proceedings of the IEEE/CVF international conference on computer vision. 2019: 9157–9166.
– reference: Lin T Y, Maire M, Belongie S, et al. Microsoft coco: Common objects in context[C]//Computer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part v 13. Springer International Publishing, 2014: 740–755.
– reference: Wang C Y, Yeh I H, Mark Liao H Y. Yolov9: Learning what you want to learn using programmable gradient information[C]//European conference on computer vision. Cham: Springer Nature Switzerland, 2024: 1–21.
– reference: Xu H, Zhang J, Cai J, et al. Gmflow: Learning optical flow via global matching. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 8121–8130.
– reference: Szegedy C, Vanhoucke V, Ioffe S, et al. Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 2818-2826.
– reference: Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020.
– reference: Wu Y, Chen Y, Yuan L, et al. Rethinking classification and localization for object detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 10186–10195.
– start-page: 1003
  year: 1996
  end-page: 1006
  ident: b0185
  article-title: Multi-scale retinex for color image enhancement
  publication-title: Proceedings of 3rd IEEE international conference on image processing
– reference: Zhang Y, Zhang J, Guo X. Kindling the darkness: A practical low-light image enhancer. In: Proceedings of the 27th ACM international conference on multimedia. 2019: 1632–1640.
– reference: Cheng B, Misra I, Schwing A G, et al. Masked-attention mask transformer for universal image segmentation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 1290–1299.
– volume: 156
  start-page: 636
  year: 1967
  end-page: 638
  ident: b0160
  article-title: How long is the coast of Britain? statistical self-similarity and fractional dimension
  publication-title: Science
– reference: Khanam R, Hussain M. Yolov11: An overview of the key architectural enhancements. arXiv preprint arXiv:2410.17725, 2024.
– reference: He K, Zhang X, Ren S, et al. Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770–778.
– start-page: 74
  year: 2024
  end-page: 90
  ident: b0110
  article-title: Rethinking features-fused-pyramid-neck for object detection
  publication-title: European Conference on Computer Vision
– reference: Kirillov A, Wu Y, He K, et al. Pointrend: Image segmentation as rendering. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 9799-9808.
– reference: Ge Z, Liu S, Wang F, et al. Yolox: Exceeding yolo series in 2021. arXiv preprint arXiv:2107.08430, 2021.
– reference: Ma N, Zhang X, Zheng H T, et al. Shufflenet v2: Practical guidelines for efficient cnn architecture design. In: Proceedings of the European conference on computer vision (ECCV). 2018: 116-131.
– reference: Guo C, Li C, Guo J, et al. Zero-reference deep curve estimation for low-light image enhancement. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 1780-1789.
– volume: 15
  start-page: 6261
  year: 2024
  ident: b0125
  article-title: Adaptative machine vision with microsecond-level accurate perception beyond human retina
  publication-title: Nature Communications
– volume: 30
  start-page: 2340
  year: 2021
  end-page: 2349
  ident: b0085
  article-title: Enlightengan: Deep light enhancement without paired supervision
  publication-title: IEEE Transactions on Image Processing
– reference: Zheng Y, Zhang M, Lu F. (2020). Optical flow in the dark. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition: 6749-6757.
– volume: 37
  start-page: 107984
  year: 2024
  end-page: 108011
  ident: b0220
  article-title: Yolov10: Real-time end-to-end object detection
  publication-title: Advances in Neural Information Processing Systems
– volume: 183
  start-page: 470
  year: 2022
  end-page: 481
  ident: b0260
  article-title: LISU: Low-light indoor scene understanding with joint learning of reflectance restoration
  publication-title: ISPRS Journal of Photogrammetry and Remote Sensing
– ident: 10.1016/j.eswa.2025.129529_b0015
  doi: 10.1109/CVPR.2018.00347
– ident: 10.1016/j.eswa.2025.129529_b0045
  doi: 10.1109/CVPRW.2017.66
– volume: 156
  start-page: 636
  issue: 3775
  year: 1967
  ident: 10.1016/j.eswa.2025.129529_b0160
  article-title: How long is the coast of Britain? statistical self-similarity and fractional dimension
  publication-title: Science
  doi: 10.1126/science.156.3775.636
– ident: 10.1016/j.eswa.2025.129529_b0190
  doi: 10.1109/CVPR.2016.91
– ident: 10.1016/j.eswa.2025.129529_b0090
– ident: 10.1016/j.eswa.2025.129529_b0065
  doi: 10.1109/CVPR42600.2020.00185
– ident: 10.1016/j.eswa.2025.129529_b0270
  doi: 10.1109/CVPR42600.2020.00678
– ident: 10.1016/j.eswa.2025.129529_b0235
  doi: 10.1109/CVPR.2019.01099
– ident: 10.1016/j.eswa.2025.129529_b0245
  doi: 10.1109/CVPR52688.2022.00795
– ident: 10.1016/j.eswa.2025.129529_b0265
  doi: 10.1109/CVPR52733.2024.01605
– ident: 10.1016/j.eswa.2025.129529_b0135
  doi: 10.1109/CVPR.2017.106
– volume: 37
  start-page: 107984
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0220
  article-title: Yolov10: Real-time end-to-end object detection
  publication-title: Advances in Neural Information Processing Systems
– volume: 30
  start-page: 2340
  year: 2021
  ident: 10.1016/j.eswa.2025.129529_b0085
  article-title: Enlightengan: Deep light enhancement without paired supervision
  publication-title: IEEE Transactions on Image Processing
  doi: 10.1109/TIP.2021.3051462
– ident: 10.1016/j.eswa.2025.129529_b0205
– ident: 10.1016/j.eswa.2025.129529_b0250
  doi: 10.1145/3343031.3350926
– volume: 183
  start-page: 470
  year: 2022
  ident: 10.1016/j.eswa.2025.129529_b0260
  article-title: LISU: Low-light indoor scene understanding with joint learning of reflectance restoration
  publication-title: ISPRS Journal of Photogrammetry and Remote Sensing
  doi: 10.1016/j.isprsjprs.2021.11.010
– ident: 10.1016/j.eswa.2025.129529_b0055
– ident: 10.1016/j.eswa.2025.129529_b0070
  doi: 10.1109/CVPR.2016.90
– volume: 21
  start-page: 62
  issue: 3
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0120
  article-title: Slim-neck by GSConv: A lightweight-design for real-time detector architectures
  publication-title: Journal of Real-Time Image Processing
  doi: 10.1007/s11554-024-01436-6
– ident: 10.1016/j.eswa.2025.129529_b0140
  doi: 10.1109/CVPR52688.2022.01167
– ident: 10.1016/j.eswa.2025.129529_b0215
– start-page: 1003
  year: 1996
  ident: 10.1016/j.eswa.2025.129529_b0185
  article-title: Multi-scale retinex for color image enhancement
– volume: 15
  start-page: 6261
  issue: 1
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0125
  article-title: Adaptative machine vision with microsecond-level accurate perception beyond human retina
  publication-title: Nature Communications
  doi: 10.1038/s41467-024-50488-6
– volume: 36
  start-page: 7050
  year: 2023
  ident: 10.1016/j.eswa.2025.129529_b0025
  article-title: Vanillanet: The power of minimalism in deep learning
  publication-title: Advances in Neural Information Processing Systems
– ident: 10.1016/j.eswa.2025.129529_b0100
  doi: 10.1109/CVPR42600.2020.00982
– year: 2008
  ident: 10.1016/j.eswa.2025.129529_b0010
– ident: 10.1016/j.eswa.2025.129529_b0210
  doi: 10.1109/CVPR.2016.308
– volume: 22
  start-page: 1761
  issue: 11
  year: 2019
  ident: 10.1016/j.eswa.2025.129529_b0195
  article-title: A deep learning framework for neuroscience
  publication-title: Nature Neuroscience
  doi: 10.1038/s41593-019-0520-2
– volume: 170
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0030
  article-title: Accurate leukocyte detection based on deformable-DETR and multi-level feature fusion for aiding diagnosis of blood diseases
  publication-title: Computers in Biology and Medicine
  doi: 10.1016/j.compbiomed.2024.107917
– ident: 10.1016/j.eswa.2025.129529_b0075
  doi: 10.1109/ICCV.2017.322
– ident: 10.1016/j.eswa.2025.129529_b0050
– ident: 10.1016/j.eswa.2025.129529_b0035
  doi: 10.1109/CVPR52688.2022.00135
– volume: 631
  start-page: 755
  issue: 8022
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0200
  article-title: AI models collapse when trained on recursively generated data
  publication-title: Nature
  doi: 10.1038/s41586-024-07566-y
– ident: 10.1016/j.eswa.2025.129529_b0155
  doi: 10.1007/978-3-030-01264-9_8
– ident: 10.1016/j.eswa.2025.129529_b0230
  doi: 10.1109/CVPRW50498.2020.00203
– volume: 36
  start-page: 50408
  year: 2023
  ident: 10.1016/j.eswa.2025.129529_b0040
  article-title: A dual-stream neural network explains the functional segregation of dorsal and ventral visual pathways in human brains
  publication-title: Advances in Neural Information Processing Systems
– volume: 131
  start-page: 2198
  issue: 8
  year: 2023
  ident: 10.1016/j.eswa.2025.129529_b0020
  article-title: Instance segmentation in the dark
  publication-title: International Journal of Computer Vision
  doi: 10.1007/s11263-023-01808-8
– start-page: 74
  year: 2024
  ident: 10.1016/j.eswa.2025.129529_b0110
  article-title: Rethinking features-fused-pyramid-neck for object detection
– ident: 10.1016/j.eswa.2025.129529_b0145
  doi: 10.1109/ICCV48922.2021.00986
– ident: 10.1016/j.eswa.2025.129529_b0005
  doi: 10.1109/ICCV.2019.00925
– volume: 22
  start-page: 1544
  issue: 10
  year: 2019
  ident: 10.1016/j.eswa.2025.129529_b0170
  article-title: Learning task-state representations
  publication-title: Nature Neuroscience
  doi: 10.1038/s41593-019-0470-8
– ident: 10.1016/j.eswa.2025.129529_b0095
– ident: 10.1016/j.eswa.2025.129529_b0255
– ident: 10.1016/j.eswa.2025.129529_b0175
  doi: 10.1109/NCC.2012.6176791
– volume: 6
  start-page: 414
  year: 1983
  ident: 10.1016/j.eswa.2025.129529_b0165
  article-title: Object vision and spatial vision: Two cortical pathways
  publication-title: Trends in neurosciences
  doi: 10.1016/0166-2236(83)90190-X
– ident: 10.1016/j.eswa.2025.129529_b0115
  doi: 10.1109/CVPR52733.2024.01440
– ident: 10.1016/j.eswa.2025.129529_b0130
  doi: 10.1007/978-3-319-10602-1_48
– ident: 10.1016/j.eswa.2025.129529_b0225
  doi: 10.1007/978-3-031-72751-1_1
– volume: 178
  start-page: 30
  year: 2019
  ident: 10.1016/j.eswa.2025.129529_b0150
  article-title: Getting to know low-light images with the exclusively dark dataset
  publication-title: Computer Vision and Image Understanding
  doi: 10.1016/j.cviu.2018.10.010
– ident: 10.1016/j.eswa.2025.129529_b0240
  doi: 10.1109/CVPR42600.2020.01020
– start-page: 3354
  year: 2012
  ident: 10.1016/j.eswa.2025.129529_b0060
  article-title: Are we ready for autonomous driving? the kitti vision benchmark suite
– ident: 10.1016/j.eswa.2025.129529_b0105
  doi: 10.1109/CVPR46437.2021.00349
– start-page: 249
  year: 2014
  ident: 10.1016/j.eswa.2025.129529_b0180
  article-title: Social learning in dogs
SSID ssj0017007
Score 2.488442
Snippet Fast and accurate object perception in low-light traffic scenes has attracted increasing attention. However, due to severe illumination degradation and the...
SourceID crossref
elsevier
SourceType Index Database
Publisher
StartPage 129529
SubjectTerms Bio-inspired vision
Deep learning
Instance segmentation
Low-light traffic
Object detection
Title A biologically inspired separable learning vision model for real-time traffic object perception in Dark
URI https://dx.doi.org/10.1016/j.eswa.2025.129529
Volume 297
WOSCitedRecordID wos001568955900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  issn: 0957-4174
  databaseCode: AIEXJ
  dateStart: 19950101
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.sciencedirect.com
  omitProxy: false
  ssIdentifier: ssj0017007
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9QwELag5cCFN6LlIR-4rVxl7Thxjisogh4qEEXsLbKTcdXVKq12t9Cf35n1Y7eAECBxiVZR7I1mPk3GM9_MMPYaKj9WfW0F-q5GlGCdMApAOFCAX6dSlb5YD5uoj4_NdNp8jIn25XqcQD0M5uqqufivqsZ7qGwqnf0LdedN8Qb-RqXjFdWO1z9S_GQUGiuR9OdU1Ue5dHQrl0BtvqlQap7CIaGwPEzDWfMN0YOcCxo3T6MjqLnE6NzNYnfjyH-hAMnbVOAzy0Q-WKxiV-hUL7eVGc-snzAim8jvOdETzN6nMwq3nN588OgyP_YVwlI7uPStjaEKmdnNKX6Wamg2hKUQiKxFOQ6zepJNloG0-5N9D6GG2QEsv1PTKKkP0F_RMWZys2_2Z9qY9kUnT9Gp9zbblbVu0PTtTj4cTo9ysqkuQlV9epFYWxVogD_-06_9ly2f5OQBuxcPE3wSQPCQ3YLhEbufBnXwaLcfs9MJ38YET5jgGRM8YYIHTPA1JjhigmdM8IgJHjDBN5jADTlh4gn78u7w5M17ESdsiE5qtRJV5VSDilKg8OA8Bgva29o3eE51BlTtS9f1TddU1kMne2OpYVtX2F5LBVJa9ZTtDOcDPGPc-cJbo4wfl7asfN9o1WkttSxdURnt99goya29CI1U2sQwnLUk5Zak3AYp7zGdRNtGVzC4eC0i4Tfr9v9x3XN2dwPYF2xntbiEl-xO9211tly8ioC5Bg_6h2Y
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=A+biologically+inspired+separable+learning+vision+model+for+real-time+traffic+object+perception+in+Dark&rft.jtitle=Expert+systems+with+applications&rft.au=Li%2C+Hulin&rft.au=Ren%2C+Qiliang&rft.au=Li%2C+Jun&rft.au=Wei%2C+Hanbing&rft.date=2026-02-01&rft.pub=Elsevier+Ltd&rft.issn=0957-4174&rft.volume=297&rft_id=info:doi/10.1016%2Fj.eswa.2025.129529&rft.externalDocID=S0957417425031446
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0957-4174&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0957-4174&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0957-4174&client=summon