On the afternoon of August 26th, Assistant Professor Song Siyang from the University of Exeter, UK, was invited to give an academic report to the faculty and students of our college. The report was presided over by Dean Mao Qirong, and teachers, graduate students, and international students in related research fields attended the event.

The theme of the report was "Research and Application of Machine Learning in Affective Computing". Affective computing is currently one of the hot research directions in the field of artificial intelligence. Based on his past research achievements, Professor Song introduced the cutting-edge research in the field of affective computing from three main aspects. Firstly, he discussed enabling machines to automatically and accurately understand the external behaviors (such as facial expressions) of target individuals through observing human visual (such as facial behavior) and audio behaviors. Secondly, he explored enabling machines to analyze the internal states (such as true personality) of target individuals through similar observations. Thirdly, he delved into endowing machines with the ability to produce human-like facial reactions, based on their accurate understanding of both internal and external behaviors of target individuals. The report focused on discussing multiple frontier machine learning technologies, such as graph representation learning algorithms with multi-dimensional edge features, invertible graph networks, and automatic neural network search.