arXiv cs.AI· Seemandhar Jain, Manmohan Chandraker·· 10 小时前AI 评分40
SplatLab：面向 3DGS 研究的自主科学发现系统
SplatLab: Autonomous Scientific Discovery for 3DGS Research
AI 导读
SplatLab 是面向 3D Gaussian Splatting（3DGS）的自主科学发现系统，将文献理解、假设生成、实现、实验精炼、证伪与论文撰写整合为统一研究流水线。
正文 · 原文
Abstract:Scientific discovery requires more than generating hypotheses: ideas must be grounded in prior work, translated into executable methods, experimentally validated, challenged by counterevidence, and communicated through reproducible findings. Despite recent advances in autonomous research agents, closing this entire loop remains challenging in rapidly evolving computer-vision fields with complex training pipelines and expensive evaluations.
We introduce SplatLab, an autonomous scientific discovery system for 3D Gaussian Splatting (3DGS) that integrates literature understanding, hypothesis generation, implementation, experimental refinement, falsification, and manuscript writing into a unified research pipeline. Given a research prompt, a target-paper URL, or no input at all, SplatLab identifies research opportunities, proposes novel methods, trains and ablates candidates across Mip-NeRF 360 and LLFF, and produces manuscripts grounded in verifiable experimental evidence.
Unlike prior systems, SplatLab performs discovery over verified baseline implementations using domain-aligned modification scaffolds. A cascade-weighted Elo tournament evolves candidate methods using measured GPU performance and verbal-gradient feedback, while a standing falsifier incorporates adversarial stress testing directly into selection. An integrity audit binds reported claims to logged experiments, and reviewer-driven feedback can initiate further experimentation rather than merely revising text.
SplatLab discovers methods achieving PSNR improvements of +0.18 to +3.78 dB over matched baselines, including +3.78 dB on astrophotographic nebula rendering, a previously unexplored setting for 3DGS. By closing the loop between scientific reasoning, executable experimentation, and evidence-backed communication, SplatLab compresses months of researcher effort into days of mostly autonomous compute.
| Comments: | Accepted at NeurIPS 2026. 14 pages, 4 figures. Project page: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.30468 [cs.CV] |
| (or arXiv:2610.30468v3 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.30468 arXiv-issued DOI via DataCite |
Submission history
From: Seemandhar Jain [view email]
[v1]
Wed, 21 Oct 2026 21:30:32 UTC (19,666 KB)
[v2]
Sun, 8 Nov 2026 04:37:54 UTC (82,078 KB)
[v3]
Tue, 10 Nov 2026 12:24:58 UTC (2,675 KB)
来源：arXiv cs.AI · arxiv.org