{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,3]],"date-time":"2026-07-03T12:26:51Z","timestamp":1783081611723,"version":"3.54.6"},"reference-count":56,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"funder":[{"DOI":"10.13039\/501100013804","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100013804","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100011246","name":"State Key Laboratory of Novel Software Technology","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100011246","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["14380079"],"award-info":[{"award-number":["14380079"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Computer Vision and Image Understanding"],"published-print":{"date-parts":[[2026,6]]},"DOI":"10.1016\/j.cviu.2026.104789","type":"journal-article","created":{"date-parts":[[2026,4,25]],"date-time":"2026-04-25T15:43:00Z","timestamp":1777131780000},"page":"104789","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["SAM-BM: An adversarial benchmark for loss functions and multi-scale objects in segment anything models"],"prefix":"10.1016","volume":"269","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-5296-6501","authenticated-orcid":false,"given":"Huanhuan","family":"Lv","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Songru","family":"Jiang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuhao","family":"Bai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tuohang","family":"Wan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-7601-7505","authenticated-orcid":false,"given":"Chang","family":"Gou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lijun","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.cviu.2026.104789_b1","unstructured":"Agnihotri,\u00a0S., Jung,\u00a0S., Keuper,\u00a0M., 2024. CosPGD: an efficient white-box adversarial attack for pixel-wise prediction tasks. In: Proceedings of the 41st International Conference on Machine Learning. pp. 416\u2013451."},{"key":"10.1016\/j.cviu.2026.104789_b2","doi-asserted-by":"crossref","DOI":"10.1016\/j.inffus.2024.102303","article-title":"Adversarial attacks and defenses in explainable artificial intelligence: A survey","author":"Baniecki","year":"2024","journal-title":"Inf. Fusion"},{"key":"10.1016\/j.cviu.2026.104789_b3","doi-asserted-by":"crossref","first-page":"25971","DOI":"10.52202\/075280-1130","article-title":"Segment anything in 3d with nerfs","volume":"36","author":"Cen","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.cviu.2026.104789_b4","article-title":"SAM-COD+: SAM-guided unified framework for weakly-supervised camouflaged object detection","author":"Chen","year":"2024","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"10.1016\/j.cviu.2026.104789_b5","doi-asserted-by":"crossref","unstructured":"Cordts,\u00a0M., Omran,\u00a0M., Ramos,\u00a0S., Rehfeld,\u00a0T., Enzweiler,\u00a0M., Benenson,\u00a0R., Franke,\u00a0U., Roth,\u00a0S., Schiele,\u00a0B., 2016. The cityscapes dataset for semantic urban scene understanding. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 3213\u20133223.","DOI":"10.1109\/CVPR.2016.350"},{"key":"10.1016\/j.cviu.2026.104789_b6","unstructured":"Croce,\u00a0F., Hein,\u00a0M., 2020. Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks. In: International Conference on Machine Learning. pp. 2206\u20132216."},{"key":"10.1016\/j.cviu.2026.104789_b7","doi-asserted-by":"crossref","unstructured":"Croce,\u00a0F., Hein,\u00a0M., 2024. Segment (Almost) Nothing: Prompt-Agnostic Adversarial Attacks on Segmentation Models. In: Conference on Secure and Trustworthy Machine Learning (SaTML). pp. 425\u2013442.","DOI":"10.1109\/SaTML59370.2024.00028"},{"key":"10.1016\/j.cviu.2026.104789_b8","unstructured":"Croce,\u00a0F., Singh,\u00a0N.D., Hein,\u00a0M., 2023. Robust Semantic Segmentation: Strong Adversarial Attacks and Fast Training of Robust Models. In: The Second Workshop on New Frontiers in Adversarial Machine Learning."},{"key":"10.1016\/j.cviu.2026.104789_b9","series-title":"European Conference on Computer Vision","article-title":"Towards reliable evaluation and fast training of robust semantic segmentation models","author":"Croce","year":"2024"},{"issue":"4","key":"10.1016\/j.cviu.2026.104789_b10","doi-asserted-by":"crossref","first-page":"2963","DOI":"10.1109\/TCSVT.2024.3361076","article-title":"Learning better video query with SAM for video instance segmentation","volume":"35","author":"Fang","year":"2025","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"10.1016\/j.cviu.2026.104789_b11","doi-asserted-by":"crossref","unstructured":"Gao,\u00a0S., Lin,\u00a0Z., Xie,\u00a0X., Zhou,\u00a0P., Cheng,\u00a0M.-M., Yan,\u00a0S., 2023. Editanything: Empowering unparalleled flexibility in image editing and generation. In: Proceedings of the 31st ACM International Conference on Multimedia. pp. 9414\u20139416.","DOI":"10.1145\/3581783.3612680"},{"key":"10.1016\/j.cviu.2026.104789_b12","unstructured":"Goodfellow,\u00a0I.J., Shlens,\u00a0J., Szegedy,\u00a0C., 2015. Explaining and harnessing adversarial examples. In: International Conference on Learning Representations."},{"key":"10.1016\/j.cviu.2026.104789_b13","series-title":"European Conference on Computer Vision","first-page":"308","article-title":"Segpgd: An effective and efficient adversarial attack for evaluating and boosting segmentation robustness","author":"Gu","year":"2022"},{"key":"10.1016\/j.cviu.2026.104789_b14","series-title":"Segment anything meets universal adversarial perturbation","author":"Han","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b15","unstructured":"Hendrycks,\u00a0D., Dietterich,\u00a0T., 2019. Benchmarking Neural Network Robustness to Common Corruptions and Perturbations. In: International Conference on Learning Representations."},{"key":"10.1016\/j.cviu.2026.104789_b16","doi-asserted-by":"crossref","DOI":"10.1016\/j.cviu.2023.103800","article-title":"Exploring the differences in adversarial robustness between vit-and CNN-based models using novel metrics","volume":"235","author":"Heo","year":"2023","journal-title":"Comput. Vis. Image Underst."},{"key":"10.1016\/j.cviu.2026.104789_b17","series-title":"On the robustness of segment anything","author":"Huang","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b18","doi-asserted-by":"crossref","DOI":"10.1016\/j.media.2023.103061","article-title":"Segment anything model for medical images?","volume":"92","author":"Huang","year":"2024","journal-title":"Med. Image Anal."},{"key":"10.1016\/j.cviu.2026.104789_b19","series-title":"IEEE International Conference on Multimedia and Expo","first-page":"1","article-title":"Cross-point adversarial attack based on feature neighborhood disruption against segment anything model","author":"Jiang","year":"2024"},{"key":"10.1016\/j.cviu.2026.104789_b20","doi-asserted-by":"crossref","first-page":"29914","DOI":"10.52202\/075280-1303","article-title":"Segment anything in high quality","volume":"36","author":"Ke","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.cviu.2026.104789_b21","doi-asserted-by":"crossref","unstructured":"Kirillov,\u00a0A., Mintun,\u00a0E., Ravi,\u00a0N., Mao,\u00a0H., Rolland,\u00a0C., Gustafson,\u00a0L., Xiao,\u00a0T., Whitehead,\u00a0S., Berg,\u00a0A.C., Lo,\u00a0W.-Y., et al., 2023. Segment anything. In: Proceedings of the IEEE\/CVF International Conference on Computer Vision. pp. 4015\u20134026.","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"10.1016\/j.cviu.2026.104789_b22","series-title":"Artificial Intelligence Safety and Security","first-page":"99","article-title":"Adversarial examples in the physical world","author":"Kurakin","year":"2018"},{"key":"10.1016\/j.cviu.2026.104789_b23","article-title":"Foodsam: Any food segmentation","author":"Lan","year":"2023","journal-title":"IEEE Trans. Multimed."},{"issue":"3","key":"10.1016\/j.cviu.2026.104789_b24","doi-asserted-by":"crossref","first-page":"279","DOI":"10.1177\/30504554251319452","article-title":"SAM-YOLO: An improved small object detection model for vehicle detection","volume":"38","author":"Liao","year":"2025","journal-title":"Eur. J. Artif. Intell."},{"issue":"2","key":"10.1016\/j.cviu.2026.104789_b25","doi-asserted-by":"crossref","first-page":"567","DOI":"10.1007\/s11263-024-02196-3","article-title":"A comprehensive study on robustness of image classification models: Benchmarking and rethinking","volume":"133","author":"Liu","year":"2025","journal-title":"Int. J. Comput. Vis."},{"key":"10.1016\/j.cviu.2026.104789_b26","article-title":"Region-guided attack on the segment anything model","author":"Liu","year":"2025","journal-title":"Neural Netw."},{"issue":"10","key":"10.1016\/j.cviu.2026.104789_b27","doi-asserted-by":"crossref","first-page":"4398","DOI":"10.1007\/s11263-024-02096-6","article-title":"Benchmarking object detection robustness against real-world corruptions","volume":"132","author":"Liu","year":"2024","journal-title":"Int. J. Comput. Vis."},{"key":"10.1016\/j.cviu.2026.104789_b28","unstructured":"Liu,\u00a0H., Wu,\u00a0Y., Yu,\u00a0Z., Vorobeychik,\u00a0Y., Zhang,\u00a0N., 2023. Slowlidar: Increasing the latency of lidar-based detection using adversarial examples. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. pp. 5146\u20135155."},{"key":"10.1016\/j.cviu.2026.104789_b29","doi-asserted-by":"crossref","DOI":"10.1016\/j.cose.2022.102847","article-title":"A survey on adversarial attacks in computer vision: Taxonomy, visualization and future directions","volume":"121","author":"Long","year":"2022","journal-title":"Comput. Secur."},{"key":"10.1016\/j.cviu.2026.104789_b30","doi-asserted-by":"crossref","unstructured":"Long,\u00a0J., Shelhamer,\u00a0E., Darrell,\u00a0T., 2015. Fully convolutional networks for semantic segmentation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 3431\u20133440.","DOI":"10.1109\/CVPR.2015.7298965"},{"key":"10.1016\/j.cviu.2026.104789_b31","article-title":"Towards deep learning models resistant to adversarial attacks","author":"Madry","year":"2018","journal-title":"Int. Conf. Learn. Represent."},{"key":"10.1016\/j.cviu.2026.104789_b32","doi-asserted-by":"crossref","unstructured":"Nesti,\u00a0F., Rossolini,\u00a0G., Nair,\u00a0S., Biondi,\u00a0A., Buttazzo,\u00a0G., 2022. Evaluating the robustness of semantic segmentation for autonomous driving against real-world adversarial patch attacks. In: Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision. pp. 2280\u20132289.","DOI":"10.1109\/WACV51458.2022.00288"},{"key":"10.1016\/j.cviu.2026.104789_b33","series-title":"Robustness of sam: Segment anything under corruptions and beyond","author":"Qiao","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b34","doi-asserted-by":"crossref","unstructured":"Ren,\u00a0G., Liu,\u00a0H., Lazarou,\u00a0M., Stathaki,\u00a0T., 2025. Multi-modal Segment Anything Model for Camouflaged Scene Segmentation. In: Proceedings of the IEEE\/CVF International Conference on Computer Vision. pp. 19882\u201319892.","DOI":"10.1109\/ICCV51701.2025.01849"},{"issue":"12","key":"10.1016\/j.cviu.2026.104789_b35","doi-asserted-by":"crossref","first-page":"18328","DOI":"10.1109\/TNNLS.2023.3314512","article-title":"On the real-world adversarial robustness of real-time semantic segmentation models for autonomous driving","volume":"35","author":"Rossolini","year":"2023","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"10.1016\/j.cviu.2026.104789_b36","series-title":"Proceedings of the AAAI Conference on Artificial Intelligence","first-page":"20470","article-title":"Tinysam: Pushing the envelope for efficient segment anything model","volume":"vol. 39","author":"Shu","year":"2025"},{"key":"10.1016\/j.cviu.2026.104789_b37","doi-asserted-by":"crossref","DOI":"10.1016\/j.cviu.2023.103913","article-title":"Towards adversarial robustness verification of no-reference image-and video-quality metrics","volume":"240","author":"Shumitskaya","year":"2024","journal-title":"Comput. Vis. Image Underst."},{"key":"10.1016\/j.cviu.2026.104789_b38","article-title":"Samrs: Scaling-up remote sensing segmentation dataset with segment anything model","volume":"36","author":"Wang","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.cviu.2026.104789_b39","doi-asserted-by":"crossref","DOI":"10.1016\/j.patcog.2024.110685","article-title":"An empirical study on the robustness of the segment anything model (sam)","author":"Wang","year":"2024","journal-title":"Pattern Recognit."},{"key":"10.1016\/j.cviu.2026.104789_b40","doi-asserted-by":"crossref","unstructured":"Wei,\u00a0X., Zhang,\u00a0R., Wu,\u00a0J., Liu,\u00a0J., Lu,\u00a0M., Guo,\u00a0Y., Zhang,\u00a0S., 2024. NTO3D: Neural Target Object 3D Reconstruction with Segment Anything. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. pp. 20352\u201320362.","DOI":"10.1109\/CVPR52733.2024.01924"},{"key":"10.1016\/j.cviu.2026.104789_b41","doi-asserted-by":"crossref","DOI":"10.1016\/j.cviu.2023.103647","article-title":"Improving the robustness of adversarial attacks using an affine-invariant gradient estimator","volume":"229","author":"Xiang","year":"2023","journal-title":"Comput. Vis. Image Underst."},{"key":"10.1016\/j.cviu.2026.104789_b42","doi-asserted-by":"crossref","unstructured":"Xiong,\u00a0Y., Varadarajan,\u00a0B., Wu,\u00a0L., Xiang,\u00a0X., Xiao,\u00a0F., Zhu,\u00a0C., Dai,\u00a0X., Wang,\u00a0D., Sun,\u00a0F., Iandola,\u00a0F., et al., 2024. Efficientsam: Leveraged masked image pretraining for efficient segment anything. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. pp. 16111\u201316121.","DOI":"10.1109\/CVPR52733.2024.01525"},{"key":"10.1016\/j.cviu.2026.104789_b43","series-title":"2024 IEEE International Automated Vehicle Validation Conference","first-page":"1","article-title":"Segment-anything models achieve zero-shot robustness in autonomous driving","author":"Yan","year":"2024"},{"key":"10.1016\/j.cviu.2026.104789_b44","doi-asserted-by":"crossref","unstructured":"Zhang,\u00a0M., Bai,\u00a0H., Zhang,\u00a0J., Zhang,\u00a0R., Wang,\u00a0C., Guo,\u00a0J., Gao,\u00a0X., 2022a. Rkformer: Runge-kutta transformer with random-connection attention for infrared small target detection. In: Proceedings of the 30th ACM International Conference on Multimedia. pp. 1730\u20131738.","DOI":"10.1145\/3503161.3547817"},{"issue":"4","key":"10.1016\/j.cviu.2026.104789_b45","doi-asserted-by":"crossref","first-page":"1511","DOI":"10.1007\/s10994-022-06251-3","article-title":"Black-box Bayesian adversarial attack with transferable priors","volume":"113","author":"Zhang","year":"2024","journal-title":"Mach. Learn."},{"key":"10.1016\/j.cviu.2026.104789_b46","series-title":"Faster segment anything: Towards lightweight sam for mobile applications","author":"Zhang","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b47","article-title":"Segment anything model for medical image segmentation: Current applications and future directions","author":"Zhang","year":"2024","journal-title":"Comput. Biol. Med."},{"key":"10.1016\/j.cviu.2026.104789_b48","series-title":"European Conference on Computer Vision","first-page":"233","article-title":"IRSAM: Advancing segment anything model for infrared small target detection","author":"Zhang","year":"2024"},{"key":"10.1016\/j.cviu.2026.104789_b49","series-title":"Proceedings of the AAAI Conference on Artificial Intelligence","first-page":"7224","article-title":"IrPruneDet: Efficient infrared small target detection via wavelet structure-regularized soft channel pruning","volume":"vol. 38","author":"Zhang","year":"2024"},{"key":"10.1016\/j.cviu.2026.104789_b50","series-title":"Attack-sam: Towards evaluating adversarial robustness of segment anything model","author":"Zhang","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b51","doi-asserted-by":"crossref","unstructured":"Zhang,\u00a0M., Zhang,\u00a0R., Yang,\u00a0Y., Bai,\u00a0H., Zhang,\u00a0J., Guo,\u00a0J., 2022b. ISNet: Shape matters for infrared small target detection. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. pp. 877\u2013886.","DOI":"10.1109\/CVPR52688.2022.00095"},{"key":"10.1016\/j.cviu.2026.104789_b52","series-title":"International Conference on Medical Image Computing and Computer-Assisted Intervention","first-page":"129","article-title":"Input augmentation with sam: Boosting medical image segmentation with segmentation foundation model","author":"Zhang","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b53","series-title":"Fast segment anything","author":"Zhao","year":"2023"},{"key":"10.1016\/j.cviu.2026.104789_b54","doi-asserted-by":"crossref","first-page":"1901","DOI":"10.1109\/TMM.2024.3521769","article-title":"Black-box targeted adversarial attack on segment anything (SAM)","volume":"27","author":"Zheng","year":"2025","journal-title":"IEEE Trans. Multimed."},{"key":"10.1016\/j.cviu.2026.104789_b55","doi-asserted-by":"crossref","first-page":"49859","DOI":"10.52202\/079017-1579","article-title":"Darksam: Fooling segment anything model to segment nothing","volume":"37","author":"Zhou","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.cviu.2026.104789_b56","article-title":"Segment everything everywhere all at once","volume":"36","author":"Zou","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."}],"container-title":["Computer Vision and Image Understanding"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1077314226001566?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1077314226001566?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,7,3]],"date-time":"2026-07-03T12:08:00Z","timestamp":1783080480000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S1077314226001566"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6]]},"references-count":56,"alternative-id":["S1077314226001566"],"URL":"https:\/\/doi.org\/10.1016\/j.cviu.2026.104789","relation":{},"ISSN":["1077-3142"],"issn-type":[{"value":"1077-3142","type":"print"}],"subject":[],"published":{"date-parts":[[2026,6]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"SAM-BM: An adversarial benchmark for loss functions and multi-scale objects in segment anything models","name":"articletitle","label":"Article Title"},{"value":"Computer Vision and Image Understanding","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.cviu.2026.104789","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}],"article-number":"104789"}}