Video Object Segmentation: Tasks, Datasets, and Methods

ISBN-13: 9783031446559 / Angielski

Ning Xu; Weiyao Lin; Xiankai Lu

Spis treści

Introduction.- VOS.- YouTubeVOS Challenges.

O autorze

Ning Xu, Ph.D., is a Research Scientist at Adobe Research. He received his Ph.D. in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign. He was an organizer of the first and second LSVOS Challenge in ECCV 2018 and ICCV 2019. He is also an organizer for the ACM MM 2020 grand challenge “Large-scale Human-centric Video Analysis in Complex Events” and ACCV 2020 tutorial “Spatial –Temporal Parsing of Objects: From Segmentation to Actions”. His research interests include image and video segmentation.

Weiyao Lin, Ph.D. is a Professor at Shanghai Jiao Tong University. He received his B.S. and M.E. from Shanghai Jiao Tong University and Ph.D. degree from the University of Washington, all in electrical engineering.

Xiankai Lu, Ph.D., is a Research Professor at Shandong University. Prior to this role, he was a research associate with Inception Institute of Artificial Intelligence at Abu Dhabi, UAE. Dr. Lu received a B. E. from the Department of Automation at Shan Dong University.

Yunchao Wei, Ph.D, is a Professor in the Center of Digital Media Information Processing at Beijing Jiaotong University. He received his Ph.D. from Beijing Jiaotong University. His current research interests include visual recognition with imperfect data, image/video segmentation and object detection, and multi-modal perception.

Blurb

This book provides a thorough overview of recent progress in video object segmentation, providing researchers and industrial practitioners with thorough information on the most important problems and developed technologies in the area. Video segmentation is a fundamental topic for video understanding in computer vision. Segmenting unique objects in a given video is useful for a variety of applications, including video conference, video editing, surveillance, and autonomous driving. Given the revolution of deep learning in computer vision problems, numerous new tasks, datasets, and methods have been recently proposed in the domain of segmentation. The book includes these recent results and findings in large-scale video object segmentation as well as benchmarks in large-scale human-centric video analysis in complex events. The authors provide readers with a comprehensive understanding of the challenges involved in video object segmentation, as well as the most effective methods for resolving them.