ActiveSplat: High-Fidelity Scene Reconstruction through Active Gaussian Splatting

2024-10-29

arXiv.org

AI Summary

This work addresses the fragmentation between online mapping, viewpoint selection, and path planning in high-fidelity scene reconstruction. We propose the first unified active perception framework integrating dense geometric representation with sparse topological structure. Methodologically: (1) we introduce a hybrid map representation—employing a sparse topological graph for efficient global path planning and viewpoint-conditioned dense Gaussian splatting for high-accuracy local reconstruction and viewpoint prediction; (2) we design a hierarchical topology-driven planning strategy that jointly optimizes trajectory diversity and local reconstruction granularity; (3) we integrate real-time Gaussian rasterization, online SLAM mapping, graph-based topological abstraction, and a viewpoint quality assessment model. Experiments across diverse scenes demonstrate significant improvements in reconstruction accuracy, data coverage, and exploration efficiency. The framework enables high-quality, photorealistic novel-view synthesis under budget constraints, achieving a balanced trade-off between scene completeness and geometric fidelity.

Technology Category

• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Computer Vision: Visual Reasoning & Symbolic Representations
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

We propose ActiveSplat, an autonomous high-fidelity reconstruction system leveraging Gaussian splatting. Taking advantage of efficient and realistic rendering, the system establishes a unified framework for online mapping, viewpoint selection, and path planning. The key to ActiveSplat is a hybrid map representation that integrates both dense information about the environment and a sparse abstraction of the workspace. Therefore, the system leverages sparse topology for efficient viewpoint sampling and path planning, while exploiting view-dependent dense prediction for viewpoint selection, facilitating efficient decision-making with promising accuracy and completeness. A hierarchical planning strategy based on the topological map is adopted to mitigate repetitive trajectories and improve local granularity given limited budgets, ensuring high-fidelity reconstruction with photorealistic view synthesis. Extensive experiments and ablation studies validate the efficacy of the proposed method in terms of reconstruction accuracy, data coverage, and exploration efficiency. Project page: https://li-yuetao.github.io/ActiveSplat/.

Problem

• Autonomous high-fidelity 3D scene reconstruction
• Unified framework for mapping and path planning
• Efficient decision-making for accurate reconstruction

Innovation

• Active Gaussian splatting for high-fidelity reconstruction
• Hybrid map representation combining dense and sparse data
• Hierarchical planning strategy for efficient trajectory optimization