May 10, 2023 Wednesday
Abstract:
In this talk, I will introduce two of our recent works at CVPR 2023. Both of them are about dataset. The first is MVImgNet (https://gaplab.cuhk.edu.cn/projects/MVImgNet/) which targets to contribute a foundation dataset for 3D vision tasks, including 3D reconstruction, 3D understanding and also 3D-aware image recognition. The second is Rabit (https://gaplab.cuhk.edu.cn/projects/RaBit/), which contributes a dataset of 3D biped cartoon characters. It contains 1,500 textured meshes, where all models are manually guaranteed topological-consistent. Based on this dataset, we also built a SMPL-like parametric method to model pose and shape and a StyleGAN-based method to model texture variance. The success of ChatGPT tells us the importance of big data and big model, I believe the era of 3D big data is coming.
Speaker Bio:
Prof. Xiaoguang HAN
Assistant Professor and President Young Scholar
The Chinese University of Hong Kong (Shenzhen)
Dr. Xiaoguang Han is now an Assistant Professor and President Young Scholar of the Chinese University of Hong Kong (Shenzhen). He received his PhD degree from the University of Hong Kong in 2017. His research interests include computer vision and computer graphics. He has published nearly 60 papers in well-known international journals and conferences, including top conferences and journals SIGGRAPH(Asia), CVPR, ICCV, ECCV, NeurIPS, ACM TOG, IEEE TPAMI, etc. He currently serves as an associate editor of Computer&Graphics and IEEE Transactions on Mobile Computing, an area chair of CVPR 2023 and NeurIPS 2023. He won the 2021 Wu Wenjun Artificial Intelligence Outstanding Youth Award, the IEEE TVCG 2021 Best Reviewer Honorable Mention, his work has won the CCF Graphics Open Source Dataset Award (DeepFashion3D), his two works were selected as the best paper candidate in CVPR 2019 and 2020.