Welcome
I have successfully defended my dissertation in June, 2019 at Department of Computer Science, University of North Carolina at Chapel Hill. My Ph.D research focus on Medical Image Analysis with Deep Learning from Sep 2015 to May 2019. In the 1st year (Aug 2014-May 2015) of my Ph.D, I work on probabilistic graphical model and RBM. From the 2nd year on, I moved to medical image analysis field. Specifically, I work on adversarial confidence learning with its application in medical image segmentation and synthesis. Another major research topic for my Ph.D defense is about low-contrast (blurry) boundary delineation. I have also participated in many other medical imaging related projects. Prior to UNC, I studied at University of Chinese Academy of Sciences in Computer Science where I researched on natural language processing (NLP), especially sentiment analysis and named entity recognition.
After graduation, I started to work in industry. In the beginning, I focused on HD mapping using deep learning and geometric based computer vision for recognition (image and point cloud) and localization (SFM) respectively. Later, I became head of perception for intelligent driving, leading a team working on 3D detection (image and point cloud), multi-object tracking, fusion and auto-labeling. I also have industry experience for large language models (LLM), including large multi-modality models (LMM). After 4.5 happy and fruitful years in my first company, I started my new journey in a second company, where I worked on video AI related projects and partially on ranking AI (Somewhat surprisingly, search, ads and recsys (搜广推) are still the dominant roles in the internet industry even after more than ten years).
I have broad interest in research and engineering works, including
computer vision oriented topics: 2D/3D detection, Image/Instance/Panoptic Segmentation, Point Cloud Understanding and vision foundation models;
natural language processing based topics: named entity recognition, sentiment analysis and embeddings;
x-modality related topics: cross-modality assisted imaging acceleration and Image/Video/Text Generation (from brain signals, e.g., fMRI/EEG/X-Ray to images/videos/text);
large model based topics: ongoing exploration, but I always believe visual will ultimately find its basic semantic units as tokens in language.
You can check my github if you are interested in some of my publications or projects.