{"title":"告密者:边界样本作为泄密水印","authors":"Alexander Schlögl, Tobias Kupek, Rainer Böhme","doi":"10.1145/3437880.3460411","DOIUrl":null,"url":null,"abstract":"Boundary samples are special inputs to artificial neural networks crafted to identify the execution environment used for inference by the resulting output label. The paper presents and evaluates algorithms to generate transparent boundary samples. Transparency refers to a small perceptual distortion of the host signal (i.e., a natural input sample). For two established image classifiers, ResNet on FMNIST and CIFAR10, we show that it is possible to generate sets of boundary samples which can identify any of four tested microarchitectures. These sets can be built to not contain any sample with a worse peak signal-to-noise ratio than 70dB. We analyze the relationship between search complexity and resulting transparency.","PeriodicalId":120300,"journal":{"name":"Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2021-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"iNNformant: Boundary Samples as Telltale Watermarks\",\"authors\":\"Alexander Schlögl, Tobias Kupek, Rainer Böhme\",\"doi\":\"10.1145/3437880.3460411\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Boundary samples are special inputs to artificial neural networks crafted to identify the execution environment used for inference by the resulting output label. The paper presents and evaluates algorithms to generate transparent boundary samples. Transparency refers to a small perceptual distortion of the host signal (i.e., a natural input sample). For two established image classifiers, ResNet on FMNIST and CIFAR10, we show that it is possible to generate sets of boundary samples which can identify any of four tested microarchitectures. These sets can be built to not contain any sample with a worse peak signal-to-noise ratio than 70dB. We analyze the relationship between search complexity and resulting transparency.\",\"PeriodicalId\":120300,\"journal\":{\"name\":\"Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2021-06-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3437880.3460411\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3437880.3460411","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
iNNformant: Boundary Samples as Telltale Watermarks
Boundary samples are special inputs to artificial neural networks crafted to identify the execution environment used for inference by the resulting output label. The paper presents and evaluates algorithms to generate transparent boundary samples. Transparency refers to a small perceptual distortion of the host signal (i.e., a natural input sample). For two established image classifiers, ResNet on FMNIST and CIFAR10, we show that it is possible to generate sets of boundary samples which can identify any of four tested microarchitectures. These sets can be built to not contain any sample with a worse peak signal-to-noise ratio than 70dB. We analyze the relationship between search complexity and resulting transparency.