{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T10:04:40Z","timestamp":1790071480352,"version":"4.0.1"},"reference-count":36,"publisher":"MDPI AG","issue":"9","license":[{"start":{"date-parts":[[2026,9,20]],"date-time":"2026-09-20T00:00:00Z","timestamp":1789862400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J. Imaging"],"abstract":"<jats:p>Object detection models can experience substantial performance degradation when the input representation at deployment differs from that used during fine-tuning. In this study, we investigate a specific grayscale-to-RGB distribution shift in which detectors fine-tuned on grayscale-derived images are subsequently evaluated on RGB inputs. Experimental observations reveal a pronounced asymmetry in this setting: RGB-trained models generally retain stable performance on grayscale inputs, whereas grayscale-trained models exhibit severe degradation and substantial run-to-run variability on RGB inputs. We analyze the chromatic information processing mechanism in the first layer of multiple YOLO-family detectors and show that grayscale fine-tuning induces chromatic variance collapse in specific channels, which is closely associated with RGB detection mismatch. Based on this analysis, we introduce ColorScore to quantify kernel chromatic sensitivity and propose a kernel-aware regularization method that controls the chromatic\u2013luminance sensitivity allocation of the first input layer, thereby mitigating variance collapse without modifying the inference architecture. Across four YOLO backbones, the proposed method reduces the mean absolute mAP50 gap between grayscale and RGB evaluation from 0.140 to 0.014. The method also alleviates performance degradation in RT-DETR-L and Faster R-CNN under the same grayscale-to-RGB evaluation setting.<\/jats:p>","DOI":"10.3390\/jimaging12090457","type":"journal-article","created":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T08:17:54Z","timestamp":1790065074000},"page":"457","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A Kernel-Aware Regularization Model for Chromatic-Robust Detection: Analysis of Grayscale-to-RGB Generalization"],"prefix":"10.3390","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-8434-5257","authenticated-orcid":false,"given":"Zehang","family":"Wang","sequence":"first","affiliation":[{"name":"College of Command and Control Engineering, Army Engineering University of the Chinese People\u2019s Liberation Army, Nanjing 210007, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"methodology","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"},{"role":"conceptualization","vocabulary":"credit"}]},{"given":"Tieyong","family":"Cao","sequence":"additional","affiliation":[{"name":"College of Command and Control Engineering, Army Engineering University of the Chinese People\u2019s Liberation Army, Nanjing 210007, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1072-5422","authenticated-orcid":false,"given":"Jibin","family":"Yang","sequence":"additional","affiliation":[{"name":"College of Command and Control Engineering, Army Engineering University of the Chinese People\u2019s Liberation Army, Nanjing 210007, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"methodology","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"given":"Zhiliang","family":"Yang","sequence":"additional","affiliation":[{"name":"College of Command and Control Engineering, Army Engineering University of the Chinese People\u2019s Liberation Army, Nanjing 210007, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"visualization","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"}]},{"given":"Kaili","family":"Qi","sequence":"additional","affiliation":[{"name":"College of Command and Control Engineering, Army Engineering University of the Chinese People\u2019s Liberation Army, Nanjing 210007, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"validation","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"}]}],"member":"1968","published-online":{"date-parts":[[2026,9,20]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"93","DOI":"10.1007\/s10462-024-11033-5","article-title":"Edge deep learning in computer vision and medical diagnostics: A comprehensive survey","volume":"58","author":"Xu","year":"2025","journal-title":"Artif. Intell. Rev."},{"key":"ref_2","unstructured":"Jocher, G., Qiu, J., Liu, M., Lyu, S., Akyon, F.C., and Kalfaoglu, M.E. (2026). Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models. arXiv."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Zhao, Y., Lv, W., Xu, S., Wei, J., Wang, G., Dang, Q., Liu, Y., and Chen, J. (2024). DETRs Beat YOLOs on Real-time Object Detection. Proceedings of the 2024 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 16\u201322 June 2024, IEEE.","DOI":"10.1109\/CVPR52733.2024.01605"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"1137","DOI":"10.1109\/TPAMI.2016.2577031","article-title":"Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks","volume":"39","author":"Ren","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"17","DOI":"10.1167\/jov.22.4.17","article-title":"Deep neural models for color classification and color constancy","volume":"22","author":"Flachot","year":"2022","journal-title":"J. Vis."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"A48","DOI":"10.1364\/JOSAA.479451","article-title":"Object-based color constancy in a deep neural network","volume":"40","author":"Gegenfurtner","year":"2023","journal-title":"J. Opt. Soc. Am. A"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Lengyel, A., Strafforello, O., Bruintjes, R.-J., Gielisse, A., and van Gemert, J. (2023). Color Equivariant Convolutional Networks. arXiv.","DOI":"10.52202\/075280-1299"},{"key":"ref_8","unstructured":"Yang, Y., O\u2019Mahony, F., and Allen-Blanchette, C. (2024). Learning Color Equivariant Representations. arXiv."},{"key":"ref_9","unstructured":"Aditya, S., Alessandro, B., and Andrea, M. (2020, January 12). Assessing The Importance of Colours For CNNs in Object Recognition. Proceedings of the NeurIPS 2020 Workshop SVRHM, Virtual."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"484","DOI":"10.1109\/TBIOM.2025.3542316","article-title":"What\u2019s Color Got to Do with It? Face Recognition in Grayscale","volume":"7","author":"Bhatta","year":"2025","journal-title":"IEEE Trans. Biom. Behav. Identity Sci."},{"key":"ref_11","first-page":"1","article-title":"LENFusion: A Joint Low-Light Enhancement and Fusion Network for Nighttime Infrared and Visible Image Fusion","volume":"73","author":"Chen","year":"2024","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"109267","DOI":"10.1016\/j.optlaseng.2025.109267","article-title":"ILLVFusion: Infrared and low-light visible image fusion based on CNN and transformer","volume":"195","author":"Xie","year":"2025","journal-title":"Opt. Lasers Eng."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Taylor, J., and Xu, Y. (2021). Joint representation of color and form in convolutional neural networks: A stimulus-rich network perspective. PLoS ONE, 16.","DOI":"10.1371\/journal.pone.0253442"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"841","DOI":"10.1002\/col.22903","article-title":"Color and luminance processing in V1 complex cells and artificial neural networks","volume":"48","author":"Bun","year":"2023","journal-title":"Color Res. Appl."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Chiu, M.C., Wang, Y., Kim, D.E.G., Chen, P.Y., and Ma, X. (2025). ColorSense: A Study on Color Vision in Machine Visual Recognition. Proceedings of the 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), Copenhagen, Denmark, 9\u201311 April 2025, IEEE.","DOI":"10.1109\/SaTML64287.2025.00043"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"89","DOI":"10.1016\/j.visres.2020.09.010","article-title":"Color for object recognition: Hue and chroma sensitivity in the deep features of convolutional neural networks","volume":"182","author":"Flachot","year":"2021","journal-title":"Vis. Res."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"7","DOI":"10.1016\/j.visres.2018.03.010","article-title":"Color encoding in biologically-inspired convolutional neural networks","volume":"151","author":"Rafegas","year":"2018","journal-title":"Vis. Res."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Sanchez-Cesteros, O., Rincon, M., Bachiller, M., and Valladares-Rodriguez, S. (2023). A Long Skip Connection for Enhanced Color Selectivity in CNN Architectures. Sensors, 23.","DOI":"10.20944\/preprints202307.1957.v1"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"858","DOI":"10.1162\/neco_a_01356","article-title":"How Convolutional Neural Network Architecture Biases Learned Opponency and Color Tuning","volume":"33","author":"Harris","year":"2021","journal-title":"Neural Comput."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"14955","DOI":"10.1523\/JNEUROSCI.4348-10.2010","article-title":"Advances in Color Science: From Retina to Behavior","volume":"30","author":"Conway","year":"2010","journal-title":"J. Neurosci. Off. J. Soc. Neurosci."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Zhang, H., and Yoshida, S. (2024). Exploring Deep Neural Networks in Simulating Human Vision through Five Optical Illusions. Appl. Sci., 14.","DOI":"10.3390\/app14083429"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Yeu, Y.H., Shapiai, M.I., Ismail, Z.H., and Fauzi, H. (2019). Investigation on Different Color Spaces on Faster RCNN for Night-Time Human Occupancy Modelling. Proceedings of the 2019 IEEE 7th Conference on Systems, Process and Control (ICSPC), Melaka, Malaysia, 13\u201314 December 2019, IEEE.","DOI":"10.1109\/ICSPC47137.2019.9068000"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1007\/s10489-026-07142-z","article-title":"To fuse or not to fuse: Enhancing military operation object detection with multimodal late fusion and color space optimization","volume":"56","author":"Dobrzycki","year":"2026","journal-title":"Appl. Intell."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"402","DOI":"10.26599\/TST.2024.9010001","article-title":"Convolutional Neural Network Image Classification Based on Different Color Spaces","volume":"30","author":"Xian","year":"2025","journal-title":"Tsinghua Sci. Technol."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Maitlo, N., Noonari, N., Ghanghro, S.A., Duraisamy, S., and Ahmed, F. (2024). Color Recognition in Challenging Lighting Environments: CNN Approach. Proceedings of the 2024 IEEE 9th International Conference for Convergence in Technology (I2CT), Pune, India, 5\u20137 April 2024, IEEE.","DOI":"10.1109\/I2CT61223.2024.10543537"},{"key":"ref_26","unstructured":"Manjunath, D., Sikdar, A., Gurunath, P., Udupa, S., and Sundaram, S. (2025). SAGA: Semantic-Aware Gray color Augmentation for Visible-to-Thermal Domain Adaptation across Multi-View Drone and Ground-Based Vision Systems. arXiv."},{"key":"ref_27","first-page":"1","article-title":"Cross-Modal Adaptation for Object Detection in Infrared Remote Sensing Imagery","volume":"22","author":"Wang","year":"2025","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Zhao, Z., Bai, H., Zhang, J., Zhang, Y., Xu, S., Lin, Z., Timofte, R., and Gool, L.V. (2023). CDDFuse: Correlation-Driven Dual-Branch Feature Decomposition for Multi-Modality Image Fusion. Proceedings of the 2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada, 17\u201324 June 2023, IEEE.","DOI":"10.1109\/CVPR52729.2023.00572"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Quan, Z., Deguchi, D., Chen, J., Zhang, C., Li, Y., Ito, S., and Murase, H. (2026). A Cross-Modal Knowledge Distillation Approach for RGB-to-Infrared Video Action Recognition. Frontiers of Computer Vision, Springer.","DOI":"10.1007\/978-981-95-4578-0_3"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Huo, F., Xu, W., Guo, J., Wang, H., and Guo, S. (2024). C2KD: Bridging the Modality Gap for Cross-Modal Knowledge Distillation. Proceedings of the 2024 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 16\u201322 June 2024, IEEE.","DOI":"10.1109\/CVPR52733.2024.01515"},{"key":"ref_31","unstructured":"Jocher, G. (2020). Ultralytics YOLOv5, Ultralytics."},{"key":"ref_32","unstructured":"Jocher, G., Chaurasia, A., and Qiu, J. (2023). Ultralytics YOLOv8, Ultralytics."},{"key":"ref_33","unstructured":"Jocher, G., and Qiu, J. (2024). Ultralytics YOLO11, Ultralytics."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"170177","DOI":"10.1038\/sdata.2017.177","article-title":"A curated mammography data set for use in computer-aided detection and diagnosis research","volume":"4","author":"Lee","year":"2017","journal-title":"Sci. Data"},{"key":"ref_35","unstructured":"Basak, S.K. (2026, September 16). Breast Cancer Medical Imaging Colorized Dataset. Available online: https:\/\/www.kaggle.com\/datasets\/shuvokumarbasakbd\/breast-cancer-medical-imaging-colorized-dataset."},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Lengyel, A., Garg, S., Milford, M., and van Gemert, J.C. Zero-Shot Day-Night Domain Adaptation with a Physics Prior. Proceedings of the 2021 IEEE\/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, 10\u201317 October 2021, IEEE.","DOI":"10.1109\/ICCV48922.2021.00436"}],"container-title":["Journal of Imaging"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2313-433X\/12\/9\/457\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T09:21:47Z","timestamp":1790068907000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2313-433X\/12\/9\/457"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,20]]},"references-count":36,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9]]}},"alternative-id":["jimaging12090457"],"URL":"https:\/\/doi.org\/10.3390\/jimaging12090457","relation":{},"ISSN":["2313-433X"],"issn-type":[{"value":"2313-433X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,20]]}}}