REVIEW 7 cited by
ActiveSplat: High-Fidelity Scene Reconstruction through Active Gaussian Splatting
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original 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 time 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. The released code will be available on our project page: https://li-yuetao.github.io/ActiveSplat/.
Forward citations
Cited by 7 Pith papers
-
DynActiveGS: Active Gaussian Splatting for Dynamic Scene Reconstruction
DynActiveGS couples uncertainty-weighted Gaussian mapping with motion-aware viewpoint and path planning, improving active 3D reconstruction in dynamic scenes.
-
IntelliCap: Intelligent Guidance for Consistent View Sampling
IntelliCap combines spatial coverage meshes with LLM-ranked spherical proxies to guide photo capture, improving 3DGS and NeRF rendering quality in indoor scenes.
-
DISCOVERSE: Efficient Robot Simulation in Complex High-Fidelity Environments
A simulation framework that combines 3D Gaussian Splatting with MuJoCo reports improved zero-shot transfer of manipulation policies from simulation to real robots.
-
GauSS-MI: Gaussian Splatting Shannon Mutual Information for Active 3D Reconstruction
GauSS-MI estimates the expected visual information gain of a viewpoint from per-Gaussian reliability probabilities, and an active reconstruction system using it achieves high visual fidelity with fewer frames.
-
PINGS: Gaussian Splatting Meets Distance Fields within a Point-Based Implicit Neural Map
PINGS jointly builds a signed distance field and a Gaussian splatting radiance field in one point-based neural map, using geometric consistency to improve both.
-
ActiveGAMER: Active GAussian Mapping through Efficient Rendering
ActiveGAMER combines 3D Gaussian Splatting mapping with a rendering-based next-best-view planner to autonomously explore and reconstruct indoor scenes with state-of-the-art accuracy and completeness on Replica and MP3...
-
ActiveGS: Active Scene Reconstruction Using Gaussian Splatting
ActiveGS combines Gaussian splatting with a voxel map and a viewpoint-distribution confidence score to guide an RGB-D robot toward poorly reconstructed areas.
Discussion (0). Continue with ORCID to comment.