Research map / 3D vision / 3D generation
Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI
Santhosh Kumar Ramakrishnan, Aaron Gokaslan, Erik Wijmans, Oleksandr Maksymets, Alexander Clegg, John M Turner, Eric Undersander, Wojciech Galuba, Andrew Westbury, Angel X Chang, Manolis Savva, Yili Zhao and 1 more
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 |
|---|---|---|---|
| Train the PointNav agent on HM3D instead of other evaluated datasets | PointNav agents with depth or RGB input · Gibson, MP3D, and HM3D test sets | Success and SPL: On MP3D (test), RGB SPL 0.68 -> 0.71; depth SPL 0.80 -> 0.83; Gibson (test) success reaches 1.00 for both sensors | Improved |
| Increase HM3D training subset from 10% to 50% to the full training set (ablation) | HM3D-trained PointNav agents · PointNav test splits | SPL: Pearson coefficient ρ = 0.88 | Improved |
| Increase training-set navigable area | PointNav agents trained on the evaluated datasets · PointNav validation sets | SPL: 0.82 to 0.97 | Improved |
| Reduce train–validation episode difficulty distribution mismatch | PointNav agents · MP3D and HM3D validation sets | SPL: -0.4 to -0.7 | Improved |
| Improve training-scene visual fidelity | PointNav agents in simulation · PointNav validation sets | SPL: -0.30 to 0.1 | No clear effect |
| Train on scenes with more reconstruction defects | PointNav agents · PointNav validation sets | SPL: -0.290 to +0.302 | Mixed |
Closest papers by meaning
- Habitat Synthetic Scenes Dataset (HSSD-200): An Analysis of 3D Scene Scale and Realism Tradeoffs for ObjectGoal Navigation (CVPR 2024 · similarity 0.88)
- AmaraSpatial-10K: A Spatially and Semantically Aligned 3D Dataset for Spatial Computing and Embodied AI (NeurIPS 2026 · similarity 0.85)
- InternScenes: A Large-scale Simulatable Indoor Scene Dataset with Realistic Layouts (NeurIPS 2025 · similarity 0.84)
- Architect: Generating Vivid and Interactive 3D Scenes with Hierarchical 2D Inpainting (NeurIPS 2024 · similarity 0.84)
- RoomTour3D: Geometry-Aware Video-Instruction Tuning for Embodied Navigation (CVPR 2025 · similarity 0.83)
- Physical AI Smart Spaces: A Large-Scale Benchmark for Multi-Camera 3D Perception in Smart Spaces (NeurIPS 2026 · similarity 0.83)
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.
