Spent two weeks in Shenzhen at our summer school "Multimodal Intelligence and AI Security" — the team I mentored won with an adversarial patch attack on a smart camera! 🏆

This September I co-organized the two-week Sino-Russian Joint Academic Summer School “Multimodal Intelligence and AI Security” at Shenzhen MSU-BIT University (SMBU), a university jointly founded by Lomonosov Moscow State University and Beijing Institute of Technology. The school brought together students from our lab and from two Chinese labs working on closely related topics.

Participants of the Multimodal Intelligence and AI Security summer school at SMBU

Week 1: lectures. The first week featured 1.5-hour lectures with lively Q&A sessions. Speakers included:

  • Alex Kot, a renowned researcher in signal processing;
  • Kede Ma, Clarivate Highly Cited Researcher, with the lecture “Compression as a Learning Principle for Visual Intelligence”;
  • Wenhan Yang, who works on image processing and extreme (AIGC) image compression.

Week 2: challenges. The second week was built around 9 research challenges tackled by 10 mixed Russian-Chinese student teams. Some challenges were taken from this year’s challenges at A* conferences, others are challenges we are preparing for A* conferences in 2027. About half of the project topics were proposed by our lab.

The overall winner was the team I mentored, which built an adversarial patch that breaks the verification system of a smart camera. Congratulations to the team! 🎉