Two papers were recently accepted at ASE’20. The first one [1] is a model-based method to estimate behavioral robustness of recurrent neural networks with Xiaoning Du, Xiaofei Xie, and Yang Liu from NTU, and Lei Ma and Jianjun Zhao from Kyushu University, Japan. The second one [2] is a search-based testing approach to detect bugs in deep learning frameworks, such as TensorFlow, CNTK, Theano, and PyTorch. This is a joint work with Qianyu Guo and Xiaohong Li from Tianjin University, China, Xiaoyu Zhang and Chao Shen from XJTU, China, and Xiaofei Xie and Yang Liu from NTU.
Preprints will follow. Stay tuned.
References
- Du, X., Li, Y., Xie, X., Ma, L., Liu, Y., & Zhao, J. (2020). MARBLE: Model-Based Robustness Analysis of Stateful Deep Learning Systems. Proceedings of the 35th IEEE/ACM International Conference on Automated Software Engineering (ASE), 423–435.
- Guo, Q., Xie, X., Li, Y., Zhang, X., Liu, Y., Li, X., & Shen, C. (2020). Audee: Automated Testing for Deep Learning Frameworks. Proceedings of the 35th IEEE/ACM International Conference on Automated Software Engineering (ASE), 486–498.