{"title":"安全的交通标志识别:针对光斑攻击的注意力通用图像涂抹机制","authors":"Hangcheng Cao, Longzhi Yuan, Guowen Xu, Ziyang He, Zhengru Fang, Yuguang Fang","doi":"arxiv-2409.04133","DOIUrl":null,"url":null,"abstract":"Traffic sign recognition systems play a crucial role in assisting drivers to\nmake informed decisions while driving. However, due to the heavy reliance on\ndeep learning technologies, particularly for future connected and autonomous\ndriving, these systems are susceptible to adversarial attacks that pose\nsignificant safety risks to both personal and public transportation. Notably,\nresearchers recently identified a new attack vector to deceive sign recognition\nsystems: projecting well-designed adversarial light patches onto traffic signs.\nIn comparison with traditional adversarial stickers or graffiti, these emerging\nlight patches exhibit heightened aggression due to their ease of implementation\nand outstanding stealthiness. To effectively counter this security threat, we\npropose a universal image inpainting mechanism, namely, SafeSign. It relies on\nattention-enabled multi-view image fusion to repair traffic signs contaminated\nby adversarial light patches, thereby ensuring the accurate sign recognition.\nHere, we initially explore the fundamental impact of malicious light patches on\nthe local and global feature spaces of authentic traffic signs. Then, we design\na binary mask-based U-Net image generation pipeline outputting diverse\ncontaminated sign patterns, to provide our image inpainting model with needed\ntraining data. Following this, we develop an attention mechanism-enabled neural\nnetwork to jointly utilize the complementary information from multi-view images\nto repair contaminated signs. Finally, extensive experiments are conducted to\nevaluate SafeSign's effectiveness in resisting potential light patch-based\nattacks, bringing an average accuracy improvement of 54.8% in three widely-used\nsign recognition models","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"26 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Secure Traffic Sign Recognition: An Attention-Enabled Universal Image Inpainting Mechanism against Light Patch Attacks\",\"authors\":\"Hangcheng Cao, Longzhi Yuan, Guowen Xu, Ziyang He, Zhengru Fang, Yuguang Fang\",\"doi\":\"arxiv-2409.04133\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Traffic sign recognition systems play a crucial role in assisting drivers to\\nmake informed decisions while driving. However, due to the heavy reliance on\\ndeep learning technologies, particularly for future connected and autonomous\\ndriving, these systems are susceptible to adversarial attacks that pose\\nsignificant safety risks to both personal and public transportation. Notably,\\nresearchers recently identified a new attack vector to deceive sign recognition\\nsystems: projecting well-designed adversarial light patches onto traffic signs.\\nIn comparison with traditional adversarial stickers or graffiti, these emerging\\nlight patches exhibit heightened aggression due to their ease of implementation\\nand outstanding stealthiness. To effectively counter this security threat, we\\npropose a universal image inpainting mechanism, namely, SafeSign. It relies on\\nattention-enabled multi-view image fusion to repair traffic signs contaminated\\nby adversarial light patches, thereby ensuring the accurate sign recognition.\\nHere, we initially explore the fundamental impact of malicious light patches on\\nthe local and global feature spaces of authentic traffic signs. Then, we design\\na binary mask-based U-Net image generation pipeline outputting diverse\\ncontaminated sign patterns, to provide our image inpainting model with needed\\ntraining data. Following this, we develop an attention mechanism-enabled neural\\nnetwork to jointly utilize the complementary information from multi-view images\\nto repair contaminated signs. Finally, extensive experiments are conducted to\\nevaluate SafeSign's effectiveness in resisting potential light patch-based\\nattacks, bringing an average accuracy improvement of 54.8% in three widely-used\\nsign recognition models\",\"PeriodicalId\":501112,\"journal\":{\"name\":\"arXiv - CS - Computers and Society\",\"volume\":\"26 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computers and Society\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.04133\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computers and Society","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.04133","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Secure Traffic Sign Recognition: An Attention-Enabled Universal Image Inpainting Mechanism against Light Patch Attacks
Traffic sign recognition systems play a crucial role in assisting drivers to
make informed decisions while driving. However, due to the heavy reliance on
deep learning technologies, particularly for future connected and autonomous
driving, these systems are susceptible to adversarial attacks that pose
significant safety risks to both personal and public transportation. Notably,
researchers recently identified a new attack vector to deceive sign recognition
systems: projecting well-designed adversarial light patches onto traffic signs.
In comparison with traditional adversarial stickers or graffiti, these emerging
light patches exhibit heightened aggression due to their ease of implementation
and outstanding stealthiness. To effectively counter this security threat, we
propose a universal image inpainting mechanism, namely, SafeSign. It relies on
attention-enabled multi-view image fusion to repair traffic signs contaminated
by adversarial light patches, thereby ensuring the accurate sign recognition.
Here, we initially explore the fundamental impact of malicious light patches on
the local and global feature spaces of authentic traffic signs. Then, we design
a binary mask-based U-Net image generation pipeline outputting diverse
contaminated sign patterns, to provide our image inpainting model with needed
training data. Following this, we develop an attention mechanism-enabled neural
network to jointly utilize the complementary information from multi-view images
to repair contaminated signs. Finally, extensive experiments are conducted to
evaluate SafeSign's effectiveness in resisting potential light patch-based
attacks, bringing an average accuracy improvement of 54.8% in three widely-used
sign recognition models