Research map / 3D vision / Face and shape
Jigsaw++: Imagining Complete Shape Priors for Object Reassembly
Jiaxin Lu, Gang Hua, Qixing Huang
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 Jigsaw++ shape prior | Jigsaw · Breaking Bad (everyday) | Chamfer distance (x10^-3): 10.5 -> 4.5 | Improved |
| add Jigsaw++ shape prior | SE(3) · Breaking Bad (everyday) | Chamfer distance (x10^-3): 22.4 -> 14.3 | Improved |
| add Jigsaw++ shape prior | SE(3) · Breaking Bad (everyday) | precision (%): 20.2 -> 37.8 | Improved |
| add Jigsaw++ shape prior | Jigsaw · Breaking Bad (everyday) | recall (%): 42.7 -> 49.5 | Improved |
| add Jigsaw++ shape prior | DGL · PartNet Lamp | Chamfer distance: 68.8 -> 46.3 | Improved |
| add Jigsaw++ shape prior | DGL · PartNet Chair | precision (%): 21.5 -> 52.0 | Improved |
| remove 20% of pieces from input | Jigsaw++ · Breaking Bad Bottle | Chamfer distance (x10^-3): 1.8 -> 2.0 | No clear effect |
| remove 20% of pieces from input | Jigsaw++ · Breaking Bad Bottle | precision (%): 61.0 -> 59.5 | No clear effect |
| use generated shape prior (GT matching) in Jigsaw's global alignment | Jigsaw · Breaking Bad fracture assembly | MAE(R) (degree): 36.3 -> 17.8 | Improved |
| add 20% noise to shape prior matching | Jigsaw++ assembly · Breaking Bad fracture assembly | MAE(R) (degree): 17.8 -> 18.2 | No clear effect |
| use retargeting instead of conditional generation fine-tuning (ablation) | Jigsaw++ · Breaking Bad (everyday) | precision (%): 46.3 -> 48.7 | Improved |
| reverse sampling steps k=1/10 versus full reverse sampling (ablation) | Jigsaw++ retargeting · Breaking Bad | reconstruction quality | Improved |
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.
- Jigsaw — Jigsaw: Learning to Assemble Multiple Fractured Objects (NeurIPS 2023) · better on 5 of 5 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
- Jigsaw: Learning to Assemble Multiple Fractured Objects (NeurIPS 2023 · similarity 0.90)
- Beyond Reassembly: Fractured Object Recovery with Missing Parts (CVPR 2026 · similarity 0.89)
- Imagine: Image-Guided 3D Part Assembly with Structure Knowledge Graph (AAAI 2025 · similarity 0.86)
- DiffAssemble: A Unified Graph-Diffusion Model for 2D and 3D Reassembly (CVPR 2024 · similarity 0.86)
- Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025 · similarity 0.85)
- PyTorchGeoNodes: Enabling Differentiable Shape Programs for 3D Shape Reconstruction (CVPR 2025 · similarity 0.85)
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.
