Research map / 3D vision / 3D generation
Semantic MapNet: Building Allocentric Semantic Maps and Representations from Egocentric Views
Vincent Cartillier, Zhile Ren, Neha Jain, Stefan Lee, Irfan Essa, Dhruv Batra
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 |
|---|---|---|---|
| project egocentric features into recurrent spatial memory, then decode labels | Semantic mapping · Matterport3D test | mIoU: 32.76 -> 36.77 | Improved |
| project egocentric features into recurrent spatial memory, then decode labels | Semantic mapping · Matterport3D test | mBF1: 33.21 -> 37.02 | Improved |
| project egocentric features into recurrent spatial memory, then decode labels | Semantic mapping · Replica | mIoU: 40.77 -> 43.12 | Improved |
| project egocentric features into recurrent spatial memory, then decode labels | Semantic mapping · Replica | mBF1: 45.86 -> 45.18 | Mixed |
| use last-layer decoder features rather than encoder features (ablation) | SMNet · Matterport3D | mIoU: 29.03 -> 36.49 | Improved |
| increase memory channels from 64 to 256 (ablation) | SMNet with last-layer features · Matterport3D | mIoU: 36.27 -> 36.49 | Improved |
| erode the egocentric semantic prediction map at full resolution (ablation) | Segment → Project · Matterport3D | mIoU: 29.34 -> 23.07 | Hurt |
| downsample egocentric semantic predictions to 120×160 (ablation) | Segment → Project without erosion · Matterport3D | mIoU: 29.34 -> 31.25 | Improved |
| apply a median filter to the projected semantic map (ablation) | Segment → Project at 120×160 · Matterport3D | mIoU: 31.25 -> 32.49 | Improved |
Closest papers by meaning
- MapNet: An Allocentric Spatial Memory for Mapping Environments (CVPR 2018 · similarity 0.89)
- Learning Navigational Visual Representations with Semantic Map Supervision (ICCV 2023 · similarity 0.89)
- Allocentric Perceiver: Disentangling Allocentric Reasoning from Egocentric Visual Priors via Frame Instantiation (NeurIPS 2026 · similarity 0.87)
- Learning Spatial Common Sense With Geometry-Aware Recurrent Networks (CVPR 2019 · similarity 0.87)
- EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI (CVPR 2024 · similarity 0.87)
- EgoThinker: Unveiling Egocentric Reasoning with Spatio-Temporal CoT (NeurIPS 2025 · similarity 0.86)
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.
