Research map / 3D vision / 3D generation
Learning to Look around Objects for Top-View Representations of Outdoor Scenes
Samuel Schulter, Menghua Zhai, Nathan Jacobs, Manmohan Chandraker
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 |
|---|---|---|---|
| add simulation and OSM supervision to the refinement CNN | BEV refinement · KITTI-Anon BEV | mean IoU: 43.73 -> 49.62 | Improved |
| add simulation and OSM supervision to the refinement CNN | BEV refinement · KITTI-Ros BEV | mean IoU: 48.82 -> 58.29 | Improved |
| add simulation and OSM supervision to the refinement CNN | BEV refinement · Cityscapes BEV | mean IoU: 34.43 -> 38.03 | Improved |
| add OSM reconstruction loss to simulation-only refinement (ablation) | BEV refinement · KITTI-Ros BEV | mean IoU: 53.08 -> 58.29 | Improved |
| replace simple refinement heuristic with learned simulation+OSM refinement | BEV refinement · Cityscapes BEV | mean IoU: 34.64 -> 38.03 | Improved |
| add hallucination step before BEV mapping (ablation) | BEV pipeline · KITTI-Anon BEV | mean IoU: 38.31 -> 43.73 | Improved |
| add hallucination step, with refinement (ablation) | BEV pipeline · KITTI-Anon BEV | mean IoU: 45.79 -> 49.62 | Improved |
| add depth prediction instead of predicting BEV directly (ablation) | BEV pipeline · KITTI-Anon BEV | mean IoU: 26.58 -> 43.73 | Improved |
| use Box+Flow warp instead of Box only (ablation) | OSM alignment · KITTI-Anon BEV | mean IoU: 47.64 -> 49.62 | Improved |
| use alignment CNN instead of LBFGS optimization (ablation) | OSM alignment · KITTI-Anon BEV | mean IoU: 25.78 -> 46.55 | Improved |
| train alignment CNN jointly with refinement module (ablation) | OSM alignment · KITTI-Anon BEV | mean IoU: 46.55 -> 49.62 | Improved |
| place boxes on any background instead of road only during mask training (ablation) | Hallucination CNN · KITTI-Anon human-gt hidden | mean IoU: 53.18 -> 60.19 | Improved |
Closest papers by meaning
- A Parametric Top-View Representation of Complex Road Scenes (CVPR 2019 · similarity 0.91)
- Weakly but Deeply Supervised Occlusion-Reasoned Parametric Road Layouts (CVPR 2022 · similarity 0.91)
- Predicting Semantic Map Representations From Images Using Pyramid Occupancy Networks (CVPR 2020 · similarity 0.88)
- SNAP: Self-Supervised Neural Maps for Visual Positioning and Semantic Understanding (NeurIPS 2023 · similarity 0.87)
- Know Your Neighbors: Improving Single-View Reconstruction via Spatial Vision-Language Reasoning (CVPR 2024 · similarity 0.86)
- RoadSceneBench: A Lightweight Benchmark for Mid-Level Road Scene Understanding (CVPR 2026 · 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.
