Adversarial ML

Your Compiler is Backdooring Your Model: Understanding and Exploiting Compilation Inconsistency Vulnerabilities in Deep Learning Compilers
Deep learning (DL) compilers are core infrastructure in modern DL systems, offering flexibility and scalability beyond vendor-specific …
Simin Chen, Jinjun Peng, Yixin He, Junfeng Yang, Baishakhi Ray
PDF Code
Dynamic Transformer Provide a False Sense of Efficiency
Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software. Create your slides in Markdown - click the Slides button to check out the example.
Yiming Chen, Simin Chen, Zexin Li, Wei Yang, Cong Liu, Robby Tan, Haizhou Li
The Dark Side of Dynamic Routing Neural Networks: Towards Efficiency Backdoor Injection
Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software. Create your slides in Markdown - click the Slides button to check out the example.
Simin Chen, Hanlin Chen, Mirazul Haque, Cong Liu, Wei Yang
Sibling-Attack: Rethinking Transferable Adversarial Attacks against Face Recognition
Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software. Create your slides in Markdown - click the Slides button to check out the example.
Zexin Li, Bangjie Yin, Taiping Yao, Junfeng Guo, Shouhong Ding, Simin Chen, Cong Liu
NICGSlowDown: Evaluating the Efficiency Robustness of Neural Caption Generation Models
Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software. Create your slides in Markdown - click the Slides button to check out the example.
Simin Chen, Zihe Song, Mirazul Haque, Cong Liu, Wei Yang
PDF Code