FreeScale：通过确定性感知的自由视角生成扩展三维场景
FreeScale: Scaling 3D Scenes via Certainty-Aware Free-View Generation
把不完美的重建场景当作几何代理，用确定性感知的自由视角采样筛选语义有意义且受重建误差影响小的新视角，从而把少量真实序列扩成可训练前馈 NVS 模型的大规模数据；这套「重建→采样→再训练」的数据引擎可直接用于缓解 3D 视觉的数据瓶颈。
原文依据 a certainty-aware free-view sampling strategy identifying novel viewpoints that are both semantically meaningful and minimally affected by reconstruction errors
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论具备泛化能力的新视角合成（NVS）模型的发展，严重受限于缺乏包含多样且精确相机轨迹的大规模训练数据。真实世界采集虽然照片级逼真，但通常稀疏且离散；合成数据虽可规模化，却存在域间差距，且往往缺乏真实语义。我们提出 FreeScale，一个利用场景重建能力将有限的真实图像序列转化为可扩展高质量训练数据来源的新框架。我们的关键洞见是：不完美的重建场景可作为丰富的几何代理，但朴素地从中采样会放大伪影。为此，我们提出一种确定性感知的自由视角采样策略，识别出既具有语义意义、又受重建误差影响最小的新视角。我们通过规模化训练前馈 NVS 模型来验证 FreeScale 的有效性，在具有挑战性的分布外基准上取得 2.7 dB 的峰值信噪比显著提升。此外，我们表明生成的数据还能主动增强逐场景三维高斯泼溅优化，在多个数据集上带来一致改进。我们的工作为克服三维视觉中的一个根本瓶颈提供了实用而强大的数据生成引擎。项目主页：https://mvp-ai-lab.github.io/FreeScale 。
英文原文摘要 1463 字符
The development of generalizable Novel View Synthesis (NVS) models is critically limited by the scarcity of large-scale training data featuring diverse and precise camera trajectories. While real-world captures are photorealistic, they are typically sparse and discrete. Conversely, synthetic data scales but suffers from a domain gap and often lacks realistic semantics. We introduce \change FreeScale , a novel framework that leverages the power of scene reconstruction to transform limited real-world image sequences into a scalable source of high-quality training data. Our key insight is that an imperfect reconstructed scene serves as a rich geometric proxy, but naively sampling from it amplifies artifacts. To this end, we propose a certainty-aware free-view sampling strategy \change identifying novel viewpoints that are both semantically meaningful and minimally affected by reconstruction errors. We demonstrate \change FreeScale 's effectiveness by scaling up the training of feedforward NVS models, achieving a \change notable gain of 2.7 dB in PSNR on challenging out-of-distribution benchmarks. Furthermore, we show that the generated data can actively enhance per-scene 3D Gaussian Splatting optimization, leading to consistent improvements across multiple datasets. Our work provides a practical and powerful data generation engine to overcome a fundamental bottleneck in 3D vision. \change Project page: https://mvp-ai-lab.github.io/FreeScale .