PuzzleFusion++: Auto-agglomerative 3D Fracture Assembly by Denoise and Verify
Zhengqing Wang, Jiacheng Chen, Yasutaka Furukawa
2025Year
8Top-tier citations
Abstract
Figure 1 : PuzzleFusion++ iteratively aligns and assembles fracture fragments into a 3D shape, resembling how humans solve jigsaw puzzles. At each iteration, a diffusion model solves for the 6-DoF alignments of the fragments, and a transformer verifies the pairwise alignments and merge them into larger fragments. We call our approach "auto-agglomerative," referring to auto-regressive methods for the iterative process and agglomeration clustering for the hierarchical grouping.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext 4b14fcc3-c8ce-4266-8ee5-d630888aff7cCited by top-tier papers8
- Rectified Point Flow: Generic Point Cloud Pose EstimationTao Sun, Liyuan Zhu, Shengyu Huang, Shuran Song et al.NeurIPS 2025 · 14 citations
- Combinative Matching for Geometric Shape AssemblyNahyuk Lee, Juhong Min, Junhong Lee, Chunghyun Park et al.ICCV 2025 · 3 citations
- GARF: Learning Generalizable 3D Reassembly for Real-World FracturesSihang Li, Zeyu Jiang, Grace Chen, Chenyang Xu et al.ICCV 2025 · 2 citations
- MEGA-Bench: Scaling Multimodal Evaluation to over 500 Real-World TasksJiacheng Chen, Tianhao Liang, Sherman Siu, Zhengqing Wang et al.ICLR 2025
- SE(n)-Invariant Flow Matching: A General Framework with Application to Object ReassemblyGaël Heck, Sylvie Le Hégarat-Mascle, Nicolas LerméICML 2026
Builds on26
- Denoising Diffusion Probabilistic ModelsJonathan Ho, Ajay Jain, Pieter AbbeelNeurIPS 2020 · 35,902 citations
- Diffusion Models Beat GANs on Image SynthesisPrafulla Dhariwal, Alexander Quinn NicholNeurIPS 2021 · 13,211 citations
- High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
- Denoising Diffusion Implicit ModelsJiaming Song, Chenlin Meng, Stefano ErmonICLR 2021 · 11,743 citations
- LION: Latent Point Diffusion Models for 3D Shape GenerationXiaohui Zeng, Arash Vahdat, Francis Williams, Zan Gojcic et al.NeurIPS 2022 · 752 citations
Related papers
- Jigsaw++: Imagining Complete Shape Priors for Object ReassemblyJiaxin Lu, Gang Hua, Qixing HuangICCV 2025 · 2 citations
- Jigsaw: Learning to Assemble Multiple Fractured ObjectsJiaxin Lu, Yifan Sun, Qixing HuangNeurIPS 2023 · 44 citations
- PuzzleFixer: A Visual Reassembly System for Immersive Fragments RestorationShuainan Ye, Chen Zhu-Tian, Xiangtong Chu, Kang Li et al.IEEE VIS 2022 · 12 citations
- Solving Masked Jigsaw Puzzles with Diffusion Vision TransformersJinyang Liu, Wondmgezahu Teshome, Sandesh Ghimire, Mario Sznaier et al.CVPR 2024
- CRAG: Can 3D Generative Models Help 3D Assembly?Zeyu Jiang, Sihang Li, Siqi Tan, Chenyang Xu et al.ICML 2026
