Research map / 3D vision / Human pose estimation
PARIS: Part-level Reconstruction and Motion Analysis for Articulated Objects
Jiayi Liu, Ali Mahdavi-Amiri, Manolis Savva
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 mobile-field regularization L_prob (one field per ray) (ablation) | PARIS · PartNet-Mobility | CD-m (movable part): 23.32 -> 5.38 | Improved |
| add L_prob regularization (ablation) | PARIS · PartNet-Mobility | PSNR: 36.64 -> 37.73 | Improved |
| use virtual canonical state t*=0.5 instead of t*=0 for the mobile field (ablation) | PARIS · PartNet-Mobility | CD-m (movable part): 74.34 -> 3.06 | Improved |
| canonical state choice for appearance quality (ablation) | PARIS · PartNet-Mobility | PSNR / SSIM: 37.62 -> 37.73 PSNR; SSIM 0.992 both | No clear effect |
| increase number of input views from 4 to 64 (ablation) | PARIS · PartNet-Mobility | CD-w and joint axis error: CD-w 170.116 (4 views) -> 3.94 (64 views) | Improved |
| image-based self-supervised PARIS instead of point-cloud Ditto for motion estimation | articulated object motion · PartNet-Mobility, 10 categories | joint axis angular error: 54.22 -> 0.069 | Improved |
| PARIS vs Ditto on whole-object reconstruction for objects with deep containers | PARIS · Fridge, Washer, Oven, Storage | CD-w: Fridge 2.68 vs Ditto 2.16; Washer 18.31 vs 7.29 | Hurt |
| PARIS on real-world scans vs Ditto for movable part geometry | PARIS · MultiScan Fridge and Storage | CD-m: Fridge 50.60 vs 67.54; Storage 20.35 vs 101.20 | Hurt |
| PARIS vs Ours-ICP (separate fields plus CSG and ICP) for static part | PARIS · PartNet-Mobility | CD-s: 4.50 vs 3.79 | No clear effect |
Later papers that compared with this method
8 later accepted papers report PARIS as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- StructureGS: Structure-aware Gaussian Splatting for Articulated Object Reconstruction (ECCV 2026 · better on 6 of 6 shared results)
- MotionAnymesh: Physics-Grounded Articulation for Simulation-Ready Digital Twins (ECCV 2026 · better on 5 of 5 shared results)
- ART: Articulated Reconstruction Transformer (CVPR 2026 · better on 3 of 3 shared results)
- ArticulatedGS: Self-supervised Digital Twin Modeling of Articulated Objects using 3D Gaussian Splatting (CVPR 2025 · better on 11 of 11 shared results)
- SplArt: Articulation Estimation and Part-Level Reconstruction with 3D Gaussian Splatting (ICCV 2025 · better on 9 of 9 shared results)
- Neural Implicit Representation for Building Digital Twins of Unknown Articulated Objects (CVPR 2024 · better on 7 of 8 shared results)
- Articulate your NeRF: Unsupervised articulated object modeling via conditional view synthesis (NeurIPS 2024 · better on 6 of 7 shared results)
- LEIA: Latent View-invariant Embeddings for Implicit 3D Articulation (ECCV 2024 · better on 3 of 6 shared results)
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.
- A-SDF — A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation (ICCV 2021) · better on 1 of 1 shared result
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
- ART: Articulated Reconstruction Transformer (CVPR 2026 · similarity 0.90)
- ArticulatedGS: Self-supervised Digital Twin Modeling of Articulated Objects using 3D Gaussian Splatting (CVPR 2025 · similarity 0.89)
- SINGAPO: Single Image Controlled Generation of Articulated Parts in Objects (ICLR 2025 · similarity 0.88)
- Articulate your NeRF: Unsupervised articulated object modeling via conditional view synthesis (NeurIPS 2024 · similarity 0.88)
- Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting (ICLR 2025 · similarity 0.88)
- Learning Implicit Representation for Reconstructing Articulated Objects (ICLR 2024 · 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.
