KPPR: Exploiting Momentum Contrast for Point Cloud-Based Place Recognition

Wiesmann, Louis; Nunes, Lucas; Behley, Jens; Stachniss, Cyrill

doi:10.1109/lra.2022.3228174

Cited by 9 publications

(2 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, the above pre-training approaches all rely on the closed-set assumption, which means that the model can barely be transferred to recognize novel categories that do not appear within the training data. Second, the above methods require accessibility to the well-registered augmented point cloud [22], [27], [32], [33] to construct the pre-training contrast views, which are very hard to obtain for large-scale 3D scenes. Third, a large number of computational power is required in the pretraining stage.…”

Section: Related Workmentioning

confidence: 99%

“…Furthermore, large-scale point cloud scenes even contain billions of points, making point-level contrastive learning intractable in computational costs. Third, the existing unsupervised contrastive learningbased pre-training for point clouds [22], [27], [32], [33] only considers geometrically registered point/voxel pairs as the positive samples, while it does not explicitly consider explicit regional information, let alone the hierarchical alignments.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

RM3D: Robust Data-Efficient 3D Scene Parsing via Traditional and Learnt 3D Descriptors-Based Semantic Region Merging

Liu

2022

Int J Comput Vis

View full text Add to dashboard Cite

Existing state-of-the-art 3D point cloud understanding methods merely perform well in a fully supervised manner. To the best of our knowledge, there exists no unified framework that simultaneously solves the downstream high-level understanding tasks including both segmentation and detection, especially when labels are extremely limited. This work presents a general and simple framework to tackle point cloud understanding when labels are limited. The first contribution is that we have done extensive methodology comparisons of traditional and learned 3D descriptors for the task of weakly supervised 3D scene understanding, and validated that our adapted traditional PFH-based 3D descriptors show excellent generalization ability across different domains. The second contribution is that we proposed a learning-based region merging strategy based on the affinity provided by both the traditional/learned 3D descriptors and learned semantics. The merging process takes both low-level geometric and high-level semantic feature correlations into consideration. Experimental results demonstrate that our framework has the best performance among the three most important weakly supervised point clouds understanding tasks including semantic segmentation, instance segmentation, and object detection even when very limited number of points are labeled. Our method, termed Region Merging 3D (RM3D), has superior performance on Scan-Net data-efficient learning online benchmarks and other four large-scale 3D understanding benchmarks under various experimental settings, outperforming current arts by a margin for various 3D understanding tasks without complicated learning strategies such as active learning.

show abstract