Research map / 3D vision / Gaussian splatting
StructureGS: Structure-aware Gaussian Splatting for Articulated Object Reconstruction
GaHye Lee, Gyoonseo Kim, Wonjong Jang, Jooeun Son, Seungyong Lee
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add structure-aware OBB losses (part fitting and contact) over photometric-only optimization (ablation) | StructureGS · PARIS | angular and Chamfer error | Improved |
| add extent regularization L_ext (ablation) | StructureGS · PARIS | CD-m | Improved |
| use structure-aware method instead of ArticulatedGS under 10 views | StructureGS · PARIS sparse views (10) | angular error (deg): 8.553 -> 0.701 | Improved |
| use structure-aware method instead of ArticulatedGS under 100 views | StructureGS · PARIS full views (100) | angular error (deg): 0.091 -> 0.088 | Improved |
| use structure-aware method instead of ArticulatedGS | StructureGS · PARIS sparse views (10) | PSNR: 23.320 -> 33.264 | Improved |
| use structure-aware method instead of ArticulatedGS | StructureGS · PARIS full views (100) | PSNR: 37.336 -> 43.39 | Improved |
| use structure-aware method instead of ScrewSplat | StructureGS · DTA Storage | CD-w (x1000): 38.15 -> 3.238 | Improved |
| use structure-aware method instead of ScrewSplat on Storage part 0 angular error | StructureGS · DTA Storage | angular error part 0 (deg): 0.052 -> 2.165 | Hurt |
| use structure-aware method instead of ScrewSplat | StructureGS · ArtGS-Multi (4-7 parts) | angular error (deg): 27.4 -> 3.67 | Improved |
| reduce training views from 100 to 10 (ablation) | ArticulatedGS · PARIS | angular error and Chamfer distance | Hurt |
| reduce training views from 100 to 10 (ablation) | StructureGS · PARIS | angular error (deg): 0.088 -> 0.701 | Hurt |
| perturb OBB initialization (clustering noise, point-cloud noise, Pi3 instead of VGGT) (ablation) | StructureGS · PARIS sparse views (10) | OBB fitting error | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PARIS — PARIS: Part-level Reconstruction and Motion Analysis for Articulated Objects (ICCV 2023) · better on 6 of 6 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
- Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting (ICLR 2025 · similarity 0.93)
- ArticulatedGS: Self-supervised Digital Twin Modeling of Articulated Objects using 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
- REArtGS: Reconstructing and Generating Articulated Objects via 3D Gaussian Splatting with Geometric and Motion Constraints (NeurIPS 2025 · similarity 0.90)
- MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.89)
- Template-free Articulated Gaussian Splatting for Real-time Reposable Dynamic View Synthesis (NeurIPS 2024 · similarity 0.89)
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.88)
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.
