mlsanityData to 7 Oct 2026

Re-assembling the past: The RePAIR dataset and benchmark for real world 2D and 3D puzzle solving

arXiv 2410.24010 ↗First posted 31 Oct 2024cs.CVcited by 17

Not on the map. Why?

Abstract (arXiv)

This paper proposes the RePAIR dataset that represents a challenging benchmark to test modern computational and data driven methods for puzzle-solving and reassembly tasks. Our dataset has unique properties that are uncommon to current benchmarks for 2D and 3D puzzle solving. The fragments and fractures are realistic, caused by a collapse of a fresco during a World War II bombing at the Pompeii archaeological park. The fragments are also eroded and have missing pieces with irregular shapes and different dimensions, challenging further the reassembly algorithms. The dataset is multi-modal providing high resolution images with characteristic pictorial elements, detailed 3D scans of the fragments and meta-data annotated by the archaeologists. Ground truth has been generated through several years of unceasing fieldwork, including the excavation and cleaning of each fragment, followed by manual puzzle solving by archaeologists of a subset of approx. 1000 pieces among the 16000 available. After digitizing all the fragments in 3D, a benchmark was prepared to challenge current reassembly and puzzle-solving methods that often solve more simplistic synthetic scenarios. The tested baselines show that there clearly exists a gap to fill in solving this computationally complex problem.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 31 of its 98 references that are in the map. First posted on arXiv 31 Oct 2024.

Builds on

31 of 98 references.

References · 31

  1. 2017Mask R-CNNcited by 7,259
  2. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  3. 2020TEASER: Fast and Certifiable Point Cloud Registrationcited by 274
  4. 2021Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentationcited by 200
  5. 2021Reviving Iterative Training with Mask Guidance for Interactive Segmentationcited by 162
  6. 2019PQ-NET: A Generative Part Seq2Seq Network for 3D Shapescited by 118
  7. 2021JoinABLe: Learning Bottom-up Assembly of Parametric CAD Jointscited by 77
  8. 2019IKEA Furniture Assembly Environment for Long-Horizon Complex Manipulation Taskscited by 68
  9. 2020Generative 3D Part Assembly via Dynamic Graph Learningcited by 65
  10. 2020Learning 3D Part Assembly from a Single Imagecited by 51
  11. 2022Neural Shape Mating: Self-Supervised Object Assembly with Adversarial Shape Priorscited by 42
  12. 2022Breaking Bad: A Dataset for Geometric Fracture and Reassemblycited by 38
19 more
  1. 2021RGL-NET: A Recurrent Graph Learning framework for Progressive Part Assemblycited by 35
  2. 2024DiffAssemble: A Unified Graph-Diffusion Model for 2D and 3D Reassemblycited by 31
  3. 2023Leveraging SE(3) Equivariance for Learning 3D Geometric Shape Assemblycited by 27
  4. 2023Jigsaw: Learning to Assemble Multiple Fractured Objectscited by 26
  5. 2020Deepzzle: Solving Visual Jigsaw Puzzles with Deep Learning andShortest Path Optimizationcited by 19
  6. 2023Fantastic Breaks: A Dataset of Paired 3D Scans of Real-World Broken Objects and Their Complete Counterpartscited by 18
  7. 2018Solving Archaeological Puzzlescited by 15
  8. 20173D Reconstruction of Incomplete Archaeological Objects Using a Generative Adversarial Networkcited by 12
  9. 2017A Genetic Algorithm-Based Solver for Very Large Jigsaw Puzzlescited by 12
  10. 2015Solving Jigsaw Puzzles with Linear Programmingcited by 9
  11. 2020Pictorial and apictorial polygonal jigsaw puzzles from arbitrary number of crossing cutscited by 9
  12. 2022GANzzle: Reframing jigsaw puzzle solving as a retrieval task using a generative mental imagecited by 9
  13. 2018JigsawNet: Shredded Image Reassembly using Convolutional Neural Network and Loop-based Compositioncited by 7
  14. 2017A Generalized Genetic Algorithm-Based Solver for Very Large Jigsaw Puzzles of Complex Typescited by 6
  15. 2022DeepMend: Learning Occupancy Functions to Represent Shape for Repaircited by 6
  16. 2022Assemble Them All: Physics-Based Planning for Generalizable Assembly by Disassemblycited by 6
  17. 2019A Novel Hybrid Scheme Using Genetic Algorithms and Deep Learning for the Reconstruction of Portuguese Tile Panelscited by 5
  18. 2022DeepJoin: Learning a Joint Occupancy, Signed Distance, and Normal Field Function for Shape Repaircited by 5
  19. 2022Batch-based Model Registration for Fast 3D Sherd Reconstructioncited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 12; 2026: 5.

17 papers.

Sort:

By area:

  1. 2025GARF: Learning Generalizable 3D Reassembly for Real-World Fracturescited by 12
  2. 2025VaseVQA: Multimodal Agent and Benchmark for Ancient Greek Potterycited by 6
  3. 2025Pairwise Alignment & Compatibility for Arbitrarily Irregular Image Fragmentscited by 5
  4. 2025BiAssemble: Learning Collaborative Affordance for Bimanual Geometric Assemblycited by 3
  5. 2025PuzLM: Solving Jigsaw Puzzles with Sequence-to-Sequence Language Modelscited by 3
  6. 2025DM-OSVP++: One-Shot View Planning Using 3D Diffusion Models for Active RGB-Based Object Reconstructioncited by 2
  7. 2025ReassembleNet: Learnable Keypoints and Diffusion for 2D Fresco Reconstructioncited by 2
  8. 2025Learning Generalizable Shape Completion with SIM(3) Equivariancecited by 2
  9. 2026The Missing GAP: From Solving Square Jigsaw Puzzles to Handling Real World Archaeological Fragmentscited by 2
  10. 2025Rejoining fragmented ancient bamboo slips with physics-driven deep learningcited by 1
  11. 2025Solving Convex Partition Visual Jigsaw Puzzlescited by 1
  12. 2025E-M3RF: An Equivariant Multimodal 3D Re-assembly Frameworkcited by 1
5 more
  1. 2026Solving Jigsaw Puzzles in the Wild: Human-Guided Reconstruction of Cultural Heritage Fragmentscited by 1
  2. 2025Benchmarking Content-Based Puzzle Solvers on Corrupted Jigsaw Puzzlescited by 0
  3. 2026Unified Primitive Proxies for Structured Shape Completioncited by 0
  4. 2026JigSync: Gauge-Resolved Synchronization for Jigsaw Reassembly under Unknown Piece Orientationcited by 0
  5. 2026Fragment-Aware Vision Transformers for Fresco-Fragment Style Classificationcited by 0