AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Clay-to-Stone: Phase-wise 3D Gaussian Splatting for Monocular Articulated Hand-Object Manipulation Modeling

Xingyu Liu, Pengfei Ren, Qi Qi, Haifeng Sun, Zirui Zhuang, Jianxin Liao, Jingyu Wang

CVPR 2026 · 3D vision › Gaussian splatting · approach: Deformable 4D Gaussian splatting

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add CLAY phase before STONE (ablation)Clay-to-Stone · ARCTIC (11 objects, per-frame articulated, average)CD: 4.22 -> 1.97Improved
add STONE phase after CLAY (ablation)Clay-to-Stone · ARCTIC (11 objects, per-frame articulated, average)CD: 2.82 -> 1.97Improved
add STONE phase after CLAY (ablation)Clay-to-Stone · ARCTIC (11 objects, per-frame articulated, average)F@5: 0.680 -> 0.741Improved
add CLAY phase (ablation)Clay-to-Stone · ARCTIC novel-view rendering (11 objects, average)PSNR: 27.63 -> 28.17Improved
add STONE phase (ablation)Clay-to-Stone · ARCTIC novel-view rendering (11 objects, average)LPIPS (x1000): 37.46 -> 35.43Improved
add CLAY or STONE phase (ablation)Clay-to-Stone · ARCTIC novel-view rendering (11 objects, average)SSIM: 0.9597 -> 0.9606No clear effect
add modulation factors beta (ablation)Clay-to-Stone · ARCTIC ablation (box, notebook, waffle iron; rendering)PSNR: 26.73 -> 28.57Improved
add visual embeddings (ablation)Clay-to-Stone · ARCTIC ablation (box, notebook, waffle iron; rendering)LPIPS (x1000): 33.18 -> 29.09Improved
add segmentation loss (SAM2) (ablation)Clay-to-Stone · ARCTIC ablation (box, notebook, waffle iron; rendering)LPIPS (x1000): 31.95 -> 29.09Improved
add temporal loss (ablation)Clay-to-Stone · ARCTIC ablation (box, notebook, waffle iron; rendering)LPIPS (x1000): 29.44 -> 29.09Improved
per-frame articulated reconstruction vs canonicalClay-to-Stone · ARCTIC vs rigid baselines (9 objects)CD: 0.79 -> 2.32Mixed
Clay-to-Stone canonical vs BIGS rigid baselineBIGS · ARCTIC vs rigid baselines (9 objects)CD: 1.28 -> 0.79Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy