SplatLab speeds up 3D Gaussian Splatting research with autonomous system
SplatLab: Autonomous Scientific Discovery for 3DGS Research
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Reproducing and extending research papers in 3D Gaussian Splatting usually takes expert engineers weeks or months. This paper presents SplatLab, a system that can automatically reproduce baseline results, propose improvements, run experiments, and write manuscripts with fully traceable claims. It uses competitions and testing to select the best ideas and ensures every reported result links back to recorded experiments. This approach cuts the research cycle from months to days while improving result quality in some cases.
What this means in practice
- •For computer vision engineers: Automate the reproduction, refinement, and evaluation of 3D scene reconstruction methods to accelerate development cycles.
- •For graphics software developers: Integrate autonomous experimentation for 3D scene modeling to improve rendering quality and reduce manual tuning effort.
Authors
Seemandhar Jain, Manmohan Chandraker
Abstract
Reproducing a published 3DGS paper takes weeks of expert engineering before any extension can be proposed. Prior autonomous-research systems read papers, draft hypotheses, and write code, but are limited by short fixed templates, low executability, and weak traceability, and none has been demonstrated end-to-end in a live computer-vision subfield with intense activity. SplatLab is an autonomous research system for 3D Gaussian Splatting that addresses these limitations. Given a research prompt, a target-paper URL, or no input at all, it reproduces a published baseline, proposes an extension, trains and ablates across Mip-NeRF 360 and LLFF, writes the manuscript, and binds every reported number to a logged training run. Five design choices distinguish it: (i) candidates mutate plugins from a verified-reproduction backbone matched to published baselines; (ii) a cascade-weighted Elo tournament across eight islands scores candidates on measured GPU PSNR with verbal-gradient feedback; (iii) a standing falsifier subtracts Elo on failure, integrating stress-test survival into selection; (iv) every reported number is bound to a logged training run, and the run cannot terminate while any claim is unbacked; (v) a seven-category 3DGS taxonomy wraps EVOLVE-BLOCK markers in domain-aligned scaffolds transferable across 3DGS repositories. On PSNR-targeted runs, SplatLab's manuscripts beat their reproduced targets by +0.18 to +3.78 dB, including +3.78 dB on astrophotographic nebula rendering, for which no prior 3DGS method exists. The research cycle compresses from months of graduate-student time to days of mostly autonomous compute. Code, all manuscripts, experiment ledgers, and the cross-run paper store are available on the project page.