TY - GEN
T1 - Path-Guided Motion Prediction with Multi-view Scene Perception
AU - Li, Zongyun
AU - Yang, Yang
AU - Gao, Xuehao
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.
PY - 2025
Y1 - 2025
N2 - Compared to the prediction for individuals, motion prediction in 3D scenes remains a challenging task, often requiring guidance on both historical motion and the surrounding environment. However, the issue of how to effectively introduce scene context into human motion prediction remains unexplored. In this paper, we propose a novel scene-aware motion prediction method that formulates the RGBD scene data through multi-view perception, while predicting the human motion that matches the scene. First, from the top view, we perform a global path planning for motion trajectory based on scene context information. Then, from the 2D view, the semantic features of the scene are extracted from image sequences and fused with the human motion features to learn the potential interaction between the scene and the motion intention. Finally, a path-guided motion prediction framework is proposed to infer the final motion of human in the 3D view. We evaluate the effectiveness of the proposed method on two challenging datasets, including both synthetic and real environments. Experimental results demonstrate that the proposed method achieves the state-of-the-art motion prediction performance in complex scenes.
AB - Compared to the prediction for individuals, motion prediction in 3D scenes remains a challenging task, often requiring guidance on both historical motion and the surrounding environment. However, the issue of how to effectively introduce scene context into human motion prediction remains unexplored. In this paper, we propose a novel scene-aware motion prediction method that formulates the RGBD scene data through multi-view perception, while predicting the human motion that matches the scene. First, from the top view, we perform a global path planning for motion trajectory based on scene context information. Then, from the 2D view, the semantic features of the scene are extracted from image sequences and fused with the human motion features to learn the potential interaction between the scene and the motion intention. Finally, a path-guided motion prediction framework is proposed to infer the final motion of human in the 3D view. We evaluate the effectiveness of the proposed method on two challenging datasets, including both synthetic and real environments. Experimental results demonstrate that the proposed method achieves the state-of-the-art motion prediction performance in complex scenes.
KW - Motion prediction
KW - Multi-view
KW - Scene-aware
UR - https://www.scopus.com/pages/publications/85209201236
U2 - 10.1007/978-981-97-8511-7_31
DO - 10.1007/978-981-97-8511-7_31
M3 - 会议稿件
AN - SCOPUS:85209201236
SN - 9789819785100
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 439
EP - 453
BT - Pattern Recognition and Computer Vision - 7th Chinese Conference, PRCV 2024, Proceedings
A2 - Lin, Zhouchen
A2 - Zha, Hongbin
A2 - Cheng, Ming-Ming
A2 - He, Ran
A2 - Liu, Cheng-Lin
A2 - Ubul, Kurban
A2 - Silamu, Wushouer
A2 - Zhou, Jie
PB - Springer Science and Business Media Deutschland GmbH
T2 - 7th Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2024
Y2 - 18 October 2024 through 20 October 2024
ER -