时间：周三（1月3日）晚上21：00-22：30
地点：八号公寓 一楼 讨论室
主题：Learning to understand visualscenes
内容：We will provide anintroduction to the research area of computer vision, focusing on its coreproblem: understanding visual scene as human brain does. In particular, we willfollow a machine learning approach to addressing a variety of challenges inscene understanding: including pixel-level, object-level and scene-levelparsing of an image or video based on algorithms that automatically learn fromdata. Many interesting vision tasks will be covered, such as objectdetection/tracking, semantic segmentation, 3D reconstruction, image synthesis,etc.
