On the afternoon of September 21, Professor Sang Jitao, the director of the Department of Computer Science of Beijing Jiaotong University, delivered an academic report at the invitation of our college. The report was conducted online and offline. The report was chaired by Professor Mao Qirong, the dean of the School of Computer Science. More than 100 teachers and students in related fields attended the meeting.

Professor Sang first clarified the topic of this lecture - the gain and loss of "non semantic features" of trusted media computing. Then, it introduces machine learning and non semantic features of "super" human, two kinds of false relevance and trusted media computing, and the unification of false relevance and non semantic feature learning. Finally, Professor Sang made some prospects for the future development of non semantic feature learning, pointing out that this field has great potential for development.

In this lecture, Professor Sang vividly introduced the advantages and disadvantages of non semantic features in the recognition process with pictures and texts. He had extensive communication and interaction with our teachers and students on related issues, deepened the understanding of the research direction and content of image recognition among the students of the School of Computer Science, and established a multi-directional information communication channel, It points out the direction for our college to further improve the professional skills and academic attainments of graduate students.