Research map / 3D vision / 3D generation
Monte Carlo Scene Search for 3D Scene Understanding
Shreyas Hampali, Sinisa Stekovic, Sayan Deb Sarkar, Chetan S. Kumar, Friedrich Fraundorfer, Vincent Lepetit
Check 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 |
|---|---|---|---|
| use the local node score instead of the simulation's global score to update Q (ablation) | MCSS · 12 challenging scenes, layout recovery | global score: 9% | Improved |
| use the local node score instead of the simulation's global score to update Q (ablation) | MCSS · 12 challenging scenes, object recovery | global score: 15% | Improved |
| use MCSS instead of random tree search | layout recovery · 12 challenging scenes | global score: 15% | Improved |
| use MCSS instead of random tree search | object recovery · 12 challenging scenes | global score: 42% | Improved |
| include the estimated layout when evaluating object hypotheses (ablation) | MCSS object retrieval · 12 challenging scenes, chair | precision: 0.58 -> 0.65 | Improved |
| include the estimated layout when evaluating object hypotheses (ablation) | MCSS object retrieval · 12 challenging scenes, chair | recall: 0.61 -> 0.84 | Improved |
| include the estimated layout when evaluating object hypotheses (ablation) | MCSS object retrieval · 12 challenging scenes, table | precision: 0.48 -> 0.66 | Improved |
| include the estimated layout when evaluating object hypotheses (ablation) | MCSS object retrieval · 12 challenging scenes, table | recall: 0.34 -> 0.58 | Improved |
| use MCSS rather than VoteNet detection with one retrieved model per box | object retrieval · Scan2CAD, chair | accuracy: 42.02 -> 74.32 | Improved |
| use MCSS rather than VoteNet detection with one retrieved model per box | object retrieval · Scan2CAD, sofa | accuracy: 27.70 -> 78.70 | Improved |
| use MCSS rather than VoteNet detection with one retrieved model per box | object retrieval · Scan2CAD, table | accuracy: 18.52 -> 24.28 | Hurt |
Closest papers by meaning
- MCTS With Refinement for Proposals Selection Games in Scene Understanding (TPAMI 2022 · similarity 0.93)
- MonteFloor: Extending MCTS for Reconstructing Accurate Large-Scale Floor Plans (ICCV 2021 · similarity 0.90)
- MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point Cloud (ECCV 2022 · similarity 0.86)
- Learning 3D Scene Priors With 2D Supervision (CVPR 2023 · similarity 0.85)
- ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning (ICLR 2025 · similarity 0.84)
- Habitat Synthetic Scenes Dataset (HSSD-200): An Analysis of 3D Scene Scale and Realism Tradeoffs for ObjectGoal Navigation (CVPR 2024 · similarity 0.84)
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.
