More about HKUST
Towards Smart, Efficient, and Reliable Semantic Image Recognition
Speaker:Dr. Dong Zhang Vision and System Design Lab The Hong Kong University of Science and Technology Title: "Towards Smart, Efficient, and Reliable Semantic Image Recognition" Date: 19 October, 2022 Time: 10:00am - 11:00am HKT Zoom link: https://hkust.zoom.us/j/465698645?pwd=aVRaNWs2RHNFcXpnWGlkR05wTTk3UT09 Meeting ID: 465 698 645 Passcode: 20222023 Abstract: Over the past years, semantic image recognition (e.g., image classification, object detection, and semantic segmentation) has achieved fantastic progress based on deep learning technologies. However, for demanding applications which require ultra-high accuracy and/or real-time performance such as autonomous driving and medical image analysis, there is still significant room for improvement. To address such challenges, a key to success is to equip the machine learning models with rich yet discriminative feature representations. In light of this context, the goal of my research strives to build smart, efficient, and reliable computer vision systems that can recognize and understand complex visual scenarios. In this talk, I will first introduce my previous contributions in these directions and their applications. Then, I will introduce some ongoing and future research plans for next-generation computer vision systems. ***************** Biography: Dr. Dong Zhang is currently a postdoctoral researcher at the Vision and System Design Lab, The Hong Kong University of Science and Technology, working with Prof. Kwang-Ting Cheng. He obtained his Ph.D. degree in Computer Science and Technology from Nanjing University of Science and Technology in Jan. 2022. From Sep. 2018 to Sep. 2020, he was supported by China Scholarship Council as a joint Ph.D. student at Nanyang Technological University. His primary research interests are in machine learning, computer vision, and medical image analysis, especially some fundamental research tasks such as image classification, semantic segmentation, object detection, and their practical applications.