Papers › Point-GCC: Universal Self-supervised 3D Scene Pre-training via Geometry-Color Contrast

Point-GCC: Universal Self-supervised 3D Scene Pre-training via Geometry-Color Contrast

31 May 2023arXiv:2305.19623archive 2025-07-28

Guofan Fan, Zekun Qi, Wenkai Shi, Kaisheng Ma

Geometry and color information provided by the point clouds are both crucial for 3D scene understanding. Two pieces of information characterize the different aspects of point clouds, but existing methods lack an elaborate design for the discrimination and relevance. Hence we explore a 3D self-supervised paradigm that can better utilize the relations of point cloud information. Specifically, we propose a universal 3D scene pre-training framework via Geometry-Color Contrast (Point-GCC), which aligns geometry and color information using a Siamese network. To take care of actual application tasks, we design (i) hierarchical supervision with point-level contrast and reconstruct and object-level contrast based on the novel deep clustering module to close the gap between pre-training and downstream tasks; (ii) architecture-agnostic backbone to adapt for various downstream models. Benefiting from the object-level representation associated with downstream tasks, Point-GCC can directly evaluate model performance and the result demonstrates the effectiveness of our methods. Transfer learning results on a wide range of tasks also show consistent improvements across all datasets. e.g., new state-of-the-art object detection results on SUN RGB-D and S3DIS datasets. Codes will be released at https://github.com/Asterisci/Point-GCC.

Paper PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

asterisci/point-gcc officialmentioned in papermentioned on GitHubpytorchMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

3D Instance Segmentation 3D Object Detection 3D Semantic Segmentation Deep Clustering Object Object Detection Scene Understanding Transfer Learning Unsupervised 3D Semantic Segmentation object-detection

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
3D Object Detection	S3DIS	Point-GCC+TR3D	mAP@0.25	75.1	#2 of 7	Archive leaderboard	report
3D Object Detection	S3DIS	Point-GCC+TR3D	mAP@0.5	56.7	#2 of 7	Archive leaderboard	report
3D Object Detection	SUN-RGBD val	Point-GCC+TR3D+FF	mAP@0.25	69.7	#1 of 32	Archive leaderboard	report
3D Object Detection	SUN-RGBD val	Point-GCC+TR3D+FF	mAP@0.5	54.0	#1 of 32	Archive leaderboard	report
3D Object Detection	SUN-RGBD val	Point-GCC+TR3D	mAP@0.25	67.7	#5 of 32	Archive leaderboard	report
3D Object Detection	SUN-RGBD val	Point-GCC+TR3D	mAP@0.5	51.0	#5 of 32	Archive leaderboard	report
3D Object Detection	ScanNetV2	Point-GCC+TR3D	mAP@0.25	73.1	#12 of 33	Archive leaderboard	report
3D Object Detection	ScanNetV2	Point-GCC+TR3D	mAP@0.5	59.6	#12 of 33	Archive leaderboard	report
Unsupervised 3D Semantic Segmentation	ScanNetV2	Point-GCC+PointNet++	mIoU	18.3	#1 of 2	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections