Deep neural networks are vulnerable to attacks from adversarial inputs.Corresponding attack research on human pose estimation(HPE),particularly for body joint detection,has been largely unexplored.Transferring classif...Deep neural networks are vulnerable to attacks from adversarial inputs.Corresponding attack research on human pose estimation(HPE),particularly for body joint detection,has been largely unexplored.Transferring classification-based attack methods to body joint regression tasks is not straightforward.Another issue is that the attack effectiveness and imperceptibility contradict each other.To solve these issues,we propose local imperceptible attacks on HPE networks.In particular,we reformulate imperceptible attacks on body joint regression into a constrained maximum allowable attack.Furthermore,we approximate the solution using iterative gradient-based strength refinement and greedy-based pixel selection.Our method crafts effective perceptual adversarial attacks that consider both human perception and attack effectiveness.We conducted a series of imperceptible attacks against state-of-the-art HPE methods,including HigherHRNet,DEKR,and ViTPose.The experimental results demonstrate that the proposed method achieves excellent imperceptibility while maintaining attack effectiveness by significantly reducing the number of perturbed pixels.Approximately 4%of the pixels can achieve sufficient attacks on HPE.展开更多
2002年,CHOW等人根据数字版权管理(Digital Rights Management,DRM)应用场景定义了白盒攻击环境的概念,并将其模型化为一种极端的攻击模型,即白盒模型。白盒模型颠覆了以往攻击模型中对攻击者能力的诸多限制,从软件保护角度考虑,攻击者...2002年,CHOW等人根据数字版权管理(Digital Rights Management,DRM)应用场景定义了白盒攻击环境的概念,并将其模型化为一种极端的攻击模型,即白盒模型。白盒模型颠覆了以往攻击模型中对攻击者能力的诸多限制,从软件保护角度考虑,攻击者被认为拥有对目标软件及其执行的完全控制权。因此,在白盒模型中,数字版权管理系统中的设备,如智能卡、机顶盒等都存在被攻击者篡改的可能。文章基于CLEFIA算法的白盒实现方案,为数字版权管理系统提供一种软件防篡改方案。该方案将软件的二进制代码文件所解释的查找表隐藏在CLEFIA算法的白盒实现方案的查找表集合中,使软件的防篡改安全性与CLEFIA算法的白盒实现方案的加解密正确性结合在一起。一旦软件发生篡改,CLEFIA算法的白盒实现方案的加解密结果将产生错误。CLEFIA算法白盒实现方案的明密文对也将发生变化,而攻击者很难对其进行修复。展开更多
In the course of reading we can guess or infer the meaning of some new words, difficult words or keywords through the information, logic, background knowledge and language structure, etc. And it is a very useful readi...In the course of reading we can guess or infer the meaning of some new words, difficult words or keywords through the information, logic, background knowledge and language structure, etc. And it is a very useful reading skill. When obtaining this kind of skill, the readers can quickly infer the meaning of new words from the clue offered by the context or the structure characteristic of new words, therefore, to improve the reading speed and reading ability.展开更多
基金National Natural Science Foundation of China,No.61972458Natural Science Foundation of Zhejiang Province,No.LZ23F020002.
文摘Deep neural networks are vulnerable to attacks from adversarial inputs.Corresponding attack research on human pose estimation(HPE),particularly for body joint detection,has been largely unexplored.Transferring classification-based attack methods to body joint regression tasks is not straightforward.Another issue is that the attack effectiveness and imperceptibility contradict each other.To solve these issues,we propose local imperceptible attacks on HPE networks.In particular,we reformulate imperceptible attacks on body joint regression into a constrained maximum allowable attack.Furthermore,we approximate the solution using iterative gradient-based strength refinement and greedy-based pixel selection.Our method crafts effective perceptual adversarial attacks that consider both human perception and attack effectiveness.We conducted a series of imperceptible attacks against state-of-the-art HPE methods,including HigherHRNet,DEKR,and ViTPose.The experimental results demonstrate that the proposed method achieves excellent imperceptibility while maintaining attack effectiveness by significantly reducing the number of perturbed pixels.Approximately 4%of the pixels can achieve sufficient attacks on HPE.
文摘2002年,CHOW等人根据数字版权管理(Digital Rights Management,DRM)应用场景定义了白盒攻击环境的概念,并将其模型化为一种极端的攻击模型,即白盒模型。白盒模型颠覆了以往攻击模型中对攻击者能力的诸多限制,从软件保护角度考虑,攻击者被认为拥有对目标软件及其执行的完全控制权。因此,在白盒模型中,数字版权管理系统中的设备,如智能卡、机顶盒等都存在被攻击者篡改的可能。文章基于CLEFIA算法的白盒实现方案,为数字版权管理系统提供一种软件防篡改方案。该方案将软件的二进制代码文件所解释的查找表隐藏在CLEFIA算法的白盒实现方案的查找表集合中,使软件的防篡改安全性与CLEFIA算法的白盒实现方案的加解密正确性结合在一起。一旦软件发生篡改,CLEFIA算法的白盒实现方案的加解密结果将产生错误。CLEFIA算法白盒实现方案的明密文对也将发生变化,而攻击者很难对其进行修复。
文摘In the course of reading we can guess or infer the meaning of some new words, difficult words or keywords through the information, logic, background knowledge and language structure, etc. And it is a very useful reading skill. When obtaining this kind of skill, the readers can quickly infer the meaning of new words from the clue offered by the context or the structure characteristic of new words, therefore, to improve the reading speed and reading ability.
