CP-VoteNet: Contrastive Prototypical VoteNet for Few-Shot Point Cloud Object Detection

TL;DR

CP-VoteNet introduces a contrastive learning approach for few-shot 3D object detection in point clouds, enhancing prototype quality by leveraging semantic and geometric relationships, leading to improved performance over existing methods.

Contribution

The paper proposes a novel contrastive learning framework that refines prototypes using semantic and geometric relationships, significantly boosting few-shot 3D object detection accuracy.

Findings

Outperforms state-of-the-art on FS-ScanNet and FS-SUNRGBD benchmarks.

Contrastive semantics mining improves categorical feature discrimination.

Primitive-level contrastive relationships enhance transferability to novel classes.

Abstract

Few-shot point cloud 3D object detection (FS3D) aims to identify and localise objects of novel classes from point clouds, using knowledge learnt from annotated base classes and novel classes with very few annotations. Thus far, this challenging task has been approached using prototype learning, but the performance remains far from satisfactory. We find that in existing methods, the prototypes are only loosely constrained and lack of fine-grained awareness of the semantic and geometrical correlation embedded within the point cloud space. To mitigate these issues, we propose to leverage the inherent contrastive relationship within the semantic and geometrical subspaces to learn more refined and generalisable prototypical representations. To this end, we first introduce contrastive semantics mining, which enables the network to extract discriminative categorical features by constructing positive and negative pairs within training batches. Meanwhile, since point features representing local patterns can be clustered into geometric components, we further propose to impose contrastive relationship at the primitive level. Through refined primitive geometric structures, the transferability of feature encoding from base to novel classes is significantly enhanced. The above designs and insights lead to our novel Contrastive Prototypical VoteNet (CP-VoteNet). Extensive experiments on two FS3D benchmarks FS-ScanNet and FS-SUNRGBD demonstrate that CP-VoteNet surpasses current state-of-the-art methods by considerable margins across different FS3D settings. Further ablation studies conducted corroborate the rationale and effectiveness of our designs.