讲座内容：Overview of Transformers on Visual Applications
讲座人：黄正能教授
讲座时间：12月27日 9:30-11:00
腾讯会议：892 2625 8886
Abstract：
The concept of attention can be modeled in machine learning as an effective weighting of data, i.e., the more informative or more important parts of data are given larger weights for the sake of more attention. Transformers are created to extract useful self-attention and/or cross-attention, which encode the input data to a hidden space and then decodes those to another domain, to generate most discriminative embedding features. Transformers, such as BERT, have been successfully used in natural language processing (NLP) to replace many existing recurrent neural networks, such as LSTM and GRU. These days, they are also shown favorable performance on visual applications, such as image classification (ViT), object detection (DETR) and multi-object tracking (Trans-Track), which are traditionally performed by convolution neural networks (CNNs). More importantly, with the feasibility of creating discriminative embedding features for both language and images, taking advantage of connecting the embedding features from both images and texts, a new trend of neural networks that can efficiently learns visual concepts from natural language supervision is thus created for zero-shot image classification or object detection. This talk will provide an overview of these technological evolutions.
Short Bio：
Dr. Jenq-Neng Hwang received the BS and MS degrees, both in electrical engineering from the National Taiwan University, Taipei, Taiwan, in 1981 and 1983 separately. He then received his Ph.D. degree from the University of Southern California. In the summer of 1989, Dr. Hwang joined the Department of Electrical and Computer Engineering (ECE) of the University of Washington in Seattle, where he has been promoted to Full Professor since 1999. He is the Director of the Information Processing Lab. (IPL), which has won several AI City Challenges and BMTT Tracking awards in the past years. Dr. Hwang served as associate editors for IEEE T-SP, T-NN and T-CSVT, T-IP and Signal Processing Magazine (SPM). He was the General Co-Chair of 2021 IEEE World AI IoT Congress, as well as the program Co-Chairs of IEEE ICME 2016, ICASSP 1998 and ISCAS 2009. Dr. Hwang is a fellow of IEEE since 2001.
