Ziyang Song

I am currently a 3rd-year PhD student at The Hong Kong Polytechnic University, supervised by Prof. Bo Yang.

Previously, I got my M.Eng and B.Eng degrees (Honors Youth Program) from Xi'an Jiaotong University. During my M.Eng study, I interned at SenseTime and Tencent Robotics X.

My general research interests lie in computer vision and machine learning. Currently I am focusing on unsupervised methods for 3D object segmentation and reconstruction.

Email / LinkedIn / Google Scholar / Github

Publications

	NVFi: Neural Velocity Fields for 3D Physics Learning from Dynamic Videos Jinxi Li, Ziyang Song, Bo Yang Advances in Neural Information Processing Systems (NeurIPS), 2023 arXiv / Code A novel representation of dynamic 3D scenes by disentangling physical velocities from geometry and appearance, enabling: 1) future frame extrapolation, 2) unsupervised semantic scene decomposition, and 3) velocity transfer.
	ActFormer: A GAN-based Transformer towards General Action-Conditioned 3D Human Motion Generation Liang Xu, Ziyang Song, Dongliang Wang, Jing Su, Zhicheng Fang, Chenjing Ding, Weihao Gan, Yichao Yan, Xin Jin, Xiaokang Yang, Wenjun Zeng, Wei Wu International Conference on Computer Vision (ICCV), 2023 arXiv / Project Page / Code (* denotes equal contribution) A GAN-based Transformer for general action-conditioned 3D human motion generation, including single-person actions and multi-person interactive actions.
	OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds Ziyang Song, Bo Yang Advances in Neural Information Processing Systems (NeurIPS), 2022 arXiv / Video / Code We propose the first unsupervised 3D object segmentation method, learning from dynamic motion patterns in point cloud sequences.
	Learning End-to-End Action Interaction by Paired-Embedding Data Augmentation Ziyang Song, Zejian Yuan, Chong Zhang, Wanchao Chi, Yonggen Ling, Shenghao Zhang Asian Conference on Computer Vision (ACCV), 2020 arXiv We specify a new task to learn end-to-end action interaction from unlabeled interactive action pairs, and a Paired-Embedding (PE) data augmentation method for efficient learning with small data.
	Attention-Oriented Action Recognition for Real-Time Human-Robot Interaction Z. Song, Ziyi Yin, Zejian Yuan, Chong Zhang, Wanchao Chi, Yonggen Ling, Shenghao Zhang International Conference on Pattern Recognition (ICPR), 2020 arXiv A framework for real-time human action recognition from RGB-D videos in human-robot interaction (HRI) scenarios.

Teaching Assistants

Fall Term, 2023: Computer Graphics (The Hong Kong Polytechnic University)

Spring Term, 2023: Creative Digital Media Design (The Hong Kong Polytechnic University)

Fall Term, 2022: Machine Learning and Data Analytics (The Hong Kong Polytechnic University)

Spring Term, 2022: Operating Systems (The Hong Kong Polytechnic University)

Fall Term, 2021: Data Structures and Database Systems (The Hong Kong Polytechnic University).

Last update: 2023.12. Thanks.