Research map / 3D vision / 3D generation
LookOut: Real-World Humanoid Egocentric Navigation
Boxiao Pan, Adam W. Harley, Francis Engelmann, C. Karen Liu, Leonidas J. Guibas
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 DINO feature encoding instead of raw RGB unprojection (ablation) | LookOut · AND held-out | L1_trans: 0.35 -> 0.17 | Improved |
| add DINO features (ablation) | LookOut · AND held-out | Col_dyn_avg: 85.3 -> 90.2 | Improved |
| add intermediate 3D feature space instead of 2D-only pooling (ablation) | LookOut · AND held-out | L1_rot: 0.44 -> 0.16 | Improved |
| add 3D feature space (ablation) | LookOut · AND held-out | Col_stt_avg: 84.9 -> 85.6 | Improved |
| add BEV projection instead of 3D convolutions (ablation) | LookOut · AND held-out | L1_trans: 0.17 -> 0.17 | No clear effect |
| add BEV projection instead of 3D convolutions (ablation) | LookOut · AND held-out | L1_rot: 0.19 -> 0.16 | Improved |
| add depth input to RGB (ablation) | LookOut · AND held-out | Col_stt_avg: 85.6 -> 87.4 | Improved |
| add depth input to RGB (ablation) | LookOut · AND held-out | L1_trans: 0.17 -> 0.15 | Improved |
| add point cloud input to RGB (ablation) | LookOut · AND held-out | Col_stt_avg: 85.6 -> 87.8 | Improved |
| use LookOut instead of EgoCast | head pose prediction · AND held-out | L1_trans: 0.34 -> 0.17 | Improved |
| use LookOut instead of EgoCast | head pose prediction · AND held-out | Col_dyn_25: 86.2 -> 90.3 | Improved |
| add goal input to LookOut | LookOut · AND held-out | L1_trans: 0.17 -> 0.11 | Improved |
Closest papers by meaning
- EgoCogNav: Cognition-aware Human Egocentric Navigation (ECCV 2026 · similarity 0.90)
- EgoTraj: Real-World Egocentric Human Trajectory (ECCV 2026 · similarity 0.90)
- Forecasting 3D Scanpaths in Egocentric Video (CVPR 2026 · similarity 0.89)
- Gaze Beyond the Frame: Forecasting Egocentric 3D Visual Span (NeurIPS 2025 · similarity 0.88)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · similarity 0.87)
- RoboSense: Large-scale Dataset and Benchmark for Egocentric Robot Perception and Navigation in Crowded and Unstructured Environments (CVPR 2025 · similarity 0.85)
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.
