ICME 2025

3D Point Cloud Analysis with Multi-modality Models

30 June - 4 July 2025 | Nantes, France

Scope

In recent years, 3D point clouds have become increasingly popular as the representation of choice for storing and manipulating 3D data, which also found numerous applications in autonomous driving, robotics, VR/AR etc. Point clouds are unordered sets of 3D coordinates which typically represent object/scene surfaces and are often captured using 3D sensors such as depth and Lidar devices. Nevertheless, its unordered, irregular, large-scale and noisy nature makes it challenging to the analysis of point cloud data. With accessing different sources, multi-modality models are easy to train, and offer opportunities for point cloud analysis tasks.

Topics of Interest 

The topics of interest include but not limited to：

• object detection

• classification 

• segmentation 

• registration 

• shape retrieval 

• compression 

• visual quality assessment

Important Dates

ICME has extended the submission deadline including Special Session

Paper Submissions：23-Dec-24  13-Dec-24

Acceptance Notification：14-Mar-25

Portal for Submission:

https://cmt3.research.microsoft.com/ICME2025

Organizers 

Xuequan Lu, The University of Western Australia, Australia

Yulan Guo, Sun Yat-sen University, China

Junhui Hou , City University of Hong Kong, SAR China