mlsanityData to 14 Oct 2026

GLEAM: Learning Generalizable Exploration Policy for Active Mapping in Complex 3D Indoor Scenes

arXiv 2505.20294 ↗First posted 26 May 2025cs.CV cs.AI cs.ROcited by 11

Not on the map. Why?

Abstract (arXiv)

Generalizable active mapping in complex unknown environments remains a critical challenge for mobile robots. Existing methods, constrained by insufficient training data and conservative exploration strategies, exhibit limited generalizability across scenes with diverse layouts and complex connectivity. To enable scalable training and reliable evaluation, we introduce GLEAM-Bench, the first large-scale benchmark designed for generalizable active mapping with 1,152 diverse 3D scenes from synthetic and real-scan datasets. Building upon this foundation, we propose GLEAM, a unified generalizable exploration policy for active mapping. Its superior generalizability comes mainly from our semantic representations, long-term navigable goals, and randomized strategies. It significantly outperforms state-of-the-art methods, achieving 66.50% coverage (+9.49%) with efficient trajectories and improved mapping accuracy on 128 unseen complex scenes. Project page: https://xiao-chen.tech/gleam/.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Builds on 42 of its 59 references that are in the map. First posted on arXiv 26 May 2025.

Builds on

42 of 59 references.

References · 42

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  3. 2017Proximal Policy Optimization Algorithmscited by 15,215
  4. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,226
  5. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  6. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  7. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,498
  8. 2017Matterport3D: Learning from RGB-D Data in Indoor Environmentscited by 1,376
  9. 2019Habitat: A Platform for Embodied AI Researchcited by 1,191
  10. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 998
  11. 2019The Replica Dataset: A Digital Replica of Indoor Spacescited by 904
  12. 2017AI2-THOR: An Interactive 3D Environment for Visual AIcited by 869
30 more
  1. 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 764
  2. 2018Open3D: A Modern Library for 3D Data Processingcited by 721
  3. 2021Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AIcited by 592
  4. 2021NICE-SLAM: Neural Implicit Scalable Encoding for SLAMcited by 589
  5. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 512
  6. 2019Leveraging Procedural Generation to Benchmark Reinforcement Learningcited by 483
  7. 2018Gibson Env: Real-World Perception for Embodied Agentscited by 481
  8. 2021iMAP: Implicit Mapping and Positioning in Real-Timecited by 476
  9. 2016Semantic Scene Completion from a Single Depth Imagecited by 472
  10. 2022ProcTHOR: Large-Scale Embodied AI Using Procedural Generationcited by 380
  11. 2020Learning to Explore using Active Neural SLAMcited by 316
  12. 2021MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learningcited by 290
  13. 2018Unity: A General Platform for Intelligent Agentscited by 236
  14. 2023Habitat Synthetic Scenes Dataset (HSSD-200): An Analysis of 3D Scene Scale and Realism Tradeoffs for ObjectGoal Navigationcited by 167
  15. 2023EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AIcited by 147
  16. 2019Learning Exploration Policies for Navigationcited by 116
  17. 2022iSDF: Real-Time Neural Signed Distance Fields for Robot Perceptioncited by 109
  18. 2020Occupancy Anticipation for Efficient Exploration and Navigationcited by 106
  19. 2021Learning Vision-Guided Quadrupedal Locomotion End-to-End with Cross-Modal Transformerscited by 79
  20. 2023FisherRF: Active View Selection and Uncertainty Quantification for Radiance Fields using Fisher Informationcited by 64
  21. 2022NeurAR: Neural Uncertainty for Autonomous 3D Reconstruction with Implicit Neural Representationscited by 55
  22. 2022Uncertainty Guided Policy for Active Robotic 3D Reconstruction using Neural Radiance Fieldscited by 54
  23. 2024GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstructioncited by 39
  24. 2022Uncertainty-driven Planner for Exploration and Navigationcited by 33
  25. 2023Active Neural Mappingcited by 31
  26. 2022ActiveRMAP: Radiance Field for Active Mapping And Planningcited by 30
  27. 2024NARUTO: Neural Active Reconstruction from Uncertain Target Observationscited by 28
  28. 2024Deep Reinforcement Learning-based Large-scale Robot Explorationcited by 16
  29. 2025NextBestPath: Efficient 3D Mapping of Unseen Environmentscited by 13
  30. 2020A Multi-Resolution Frontier-Based Planner for Autonomous 3D Explorationcited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 5; 2026: 6.

11 papers.

  1. 2025InfiniBench: Infinite Benchmarking for Visual Spatial Reasoning with Customizable Scene Complexitycited by 5
  2. 2026MAGICIAN: Efficient Long-Term Planning with Imagined Gaussians for Active Mappingcited by 5
  3. 2025AREA3D: Active Reconstruction Agent with Unified Feed-Forward 3D Perception and Vision-Language Guidancecited by 4
  4. 2025Video4Spatial: Towards Visuospatial Intelligence with Context-Guided Video Generationcited by 3
  5. 2026Explore Like Humans: Autonomous Exploration with Online SG-Memo Construction for Embodied Agentscited by 3
  6. 2025SEA: Semantic Map Prediction for Active Exploration of Uncertain Areascited by 2
  7. 2026ObjSplat: Geometry-Aware Gaussian Surfels for Active Object Reconstructioncited by 2
  8. 2025Hestia: Voxel-Face-Aware Hierarchical Next-Best-View Acquisition for Efficient 3D Reconstructioncited by 1
  9. 2026Remember to be Curious: Episodic Context and Persistent Worlds for 3D Explorationcited by 0
  10. 2026VANDERER: Map-Free Exploration using Future-Aware and Visual-Curiosity-Guided Diffusion Policycited by 0
  11. 2026Multi-Scale Semantic Mapping in Urban Environments via Observation Calibration and Policy Dependence Regularizationcited by 0