Graph-Guided Dual-Level Augmentation for 3D Scene Segmentation

2025-07-30

AI Summary

Existing 3D point cloud segmentation data augmentation methods primarily focus on local geometric transformations or semantic recombination, neglecting global structural dependencies among scene-level objects. To address this, we propose a graph-guided dual-level augmentation framework: first, learning object relationships from real point clouds to construct a structure-guiding graph; second, jointly enforcing local geometric-semantic constraints and global topological alignment to generate high-fidelity synthetic scenes that preserve both structural integrity and semantic consistency. This work is the first to explicitly incorporate graph-structured modeling into 3D point cloud data augmentation, balancing local geometric realism with global scene plausibility. Evaluated on ScanNet, S3DIS, and SemanticKITTI benchmarks, our augmented data consistently improves the generalization performance of diverse segmentation models, demonstrating the effectiveness and cross-dataset applicability of structure-aware augmentation.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

3D point cloud segmentation aims to assign semantic labels to individual points in a scene for fine-grained spatial understanding. Existing methods typically adopt data augmentation to alleviate the burden of large-scale annotation. However, most augmentation strategies only focus on local transformations or semantic recomposition, lacking the consideration of global structural dependencies within scenes. To address this limitation, we propose a graph-guided data augmentation framework with dual-level constraints for realistic 3D scene synthesis. Our method learns object relationship statistics from real-world data to construct guiding graphs for scene generation. Local-level constraints enforce geometric plausibility and semantic consistency between objects, while global-level constraints maintain the topological structure of the scene by aligning the generated layout with the guiding graph. Extensive experiments on indoor and outdoor datasets demonstrate that our framework generates diverse and high-quality augmented scenes, leading to consistent improvements in point cloud segmentation performance across various models.

Problem

• Enhance 3D point cloud segmentation via structural augmentation
• Address lack of global structural dependencies in augmentation
• Improve scene synthesis with dual-level geometric constraints

Innovation

• Graph-guided framework for 3D scene synthesis
• Dual-level constraints ensure geometric plausibility
• Aligns generated layout with global topology