The Association for the Advancement of Artificial Intelligence (AAAI) is an international scientific society devoted to promote research in, and responsible use of, artificial intelligence. AAAI Conference on Artificial Intelligence is one of the top conferences in the field of artificial intelligence and machine learning. The AAAI-23 is one of the toughest conferences in AI community with an acceptance rate of only 19.6%.

The UNNC supervision team includes Dr Jianfeng Ren, who serves as Assistant Professor of the Department of Computer Science, UNNC, and has published more than 40 papers on reputable journals and conferences including IEEE Transactions on Image Processing and Pattern Recognition, and Professor Ruibin Bai, who serves as Full Professor and the head of the school of computer science, UNNC. Prof Bai serves as several Editorial boards in international SCI journals such as EJOR and Networks. He has published more than 100 papers on journals and conferences including INFORMS Journal on Computing, Transportation Research Part B: Methodological, IEEE Transactions on Evolutionary Computation, Expert Systems with Applications, European Journal of Operational Research.

The corresponding author of both papers, Dr Jianfeng Ren, said, “I am glad to see two research outcomes in our Lab published in the top conference. Our research on vision understanding and vision reasoning supports the development of these areas. I am looking forward to seeing more achievements presented in the top journals and conferences.”

The paper “Siamese Discriminant Deep Reinforcement Learning for Solving Jigsaw Puzzles with Large Eroded Gaps”, co-authored by 4 PhD students Xingke Song, Jiahuan Jin, Chenglin Yao, Shihe Wang and two supervisors Dr Jianfeng Ren and Professor Ruibin Bai, proposed a framework integrating both computer vision and deep reinforcement learning for solving jigsaw puzzles. Xingke Song and Chenglin Yao gave an oral presentation and illustrated its application to the restoration and protection of cultural relics.

The four student authors of this paper, Xingke Song, Jiahuan Jin, Chenglin Yao and Shihe Wang, all obtained Bachelor’s degrees from the UNNC Department of Computer Science and are now studying for the PhD programme here. Xingke said that the open and initiative teaching philosophy of the UNNC deeply inspired him. The supervision team's dedication to research is also a model for all UNNCers.

The authors of the paper “Hierarchical ConViT with Attention-based Relational Reasoner for Visual Analogical Reasoning” proposed a solution model based on convolution and vision transformer, and utilize the attention mechanism to extract visual clues about analogical reasoning problems, and hence evaluated it on the core problem of analogical reasoning – “Raven’s Progressive Matrices”, as known as human IQ tests. A significant and consistent performance improvement is achieved, and the proposed method achieved the state-of-the-art results on multiple datasets, which increases the generalization ability of the intelligent system.

The first author of the paper, Wentao He, studied at the Imperial College after graduating from the UNNC. Jialu Zhang is also a graduate from the Department of Computer Science, UNNC. Wentao told us that the UNNC has a conducive research environment and an inspiring supervision team. He has a chance to work with the top scholars in the world.