The laboratory's academic paper "Advanced Attack Type I: Cheat Classifiers By Significant Changes" was accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence, the top journal in the field of machine learning. The authors are Sanli Tang, Mingjian Chen and Chengjin Sun, postgraduates of the laboratory, and the corresponding authors are Professor Jie Yang and teacher Xiaolin Huang.

      This paper proposes another type of adversarial attack that can cheat classifiers by significant changes. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Experimental results show that this paper's method is practical and effective to generate Type I adversarial examples on large-scale image datasets.