Publications

Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Min Zhang, Taulue Chen, Jun Sun

'QVIP: An ILP-based Formal Verification Approach for Quantized Neural Networks'

ASE 2022, CCF-A

Zhe Zhao, Yedi Zhang, Guangke Chen, Fu Song, Taolue Chen and Jiaxiang Liu

'ACROBAT: Accelerating CEGAR-based Neural Network Verification via Adversarial Attacks'

SAS 2022, CCF-B

Guangke Chen, Zhe Zhao, Fu Song, Sen Chen, Lingling Fan, Yang Liu

'AS2T: Arbitrary source-to-target adversarial attack on speaker recognition systems'

IEEE Transactions on Dependable and Secure Computing (TDSC), CCF-A, IF=6.791

[paper] [blog]

Guangke Chen, Zhe Zhao, Fu Song, Sen Chen, Lingling Fan, Feng Wang, Jiashui Wang

'Towards Understanding and Mitigating Audio Adversarial Examples for Speaker Recognition'

Submitted to TDSC

[paper] [code] [website]

Zhe Zhao, Guangke Chen, Tong Liu, Taishan Li, Fu Song, Jingyi Wang, Jun Sun

'Turn Lemons into Lemonade: Using Adversarial Attack Methods to Detect Abnormal Examples'

Under major revision of TDSC

Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Taolue Chen

'Precise Quantitative Analysis of Binarized Neural Networks: A BDD-based Approach'

Under major revision of TOSEM

Guangke Chen, Zhe Zhao, Fu Song, Sen Chen, Lingling Fan, Yang Liu.

'SEC4SR: A Security Analysis Platform for Speaker Recognition'.

Preprint

[paper] [code] [website]

Guangke Chen, Sen Chen, Lingling Fan, Xiaoning Du, Zhe Zhao, Fu Song, Yang Liu.

'Who is Real Bob? Adversarial Attacks on Speaker Recognition Systems'.

Oakland 2021, CCF-A

Acceptance Rate: 115/952=12%

[paper] [website] [code] [slide] [talk] [blog] [press-1] [press-2] [press-3]

Zhe Zhao, Guangke Chen, Jingyi Wang, Yiwei Yang, Fu Song, Jun Sun.

'Attack as Defense: Characterizing Adversarial Examples using Robustness'.

ISSTA 2021, CCF-A

Acceptance Rate: 51/219=23%

[paper] [code]

Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Taolue Chen.

'BDD4BNN: A BDD-based Quantitative Analysis Framework for Binarized Neural Networks'.

CAV 2021, CCF-A

Acceptance Rate: 79/290=27%

[paper]