Research map / Vision-language and multimodal / Robot action and navigation
ORION: Ordinal Neural Collapse as a Representation Prior for Visual Navigation
E In Son, Jung-Taak Kim, Seung-Woo Seo
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 ORION ordinal-collapse encoder pretraining | NoMaD · Citysim Multi-Intersection | success rate (%): 48 -> 74 | Improved |
| add ORION pretraining | NoMaD · Indoor 2D-3D-S Intersection | success rate (%): 17 -> 91 | Improved |
| add ORION pretraining | NoMaD · Citysim Multi-Intersection | heading jerk: 34.49 -> 10.13 | Improved |
| add ORION pretraining | NoMaD · Real-world Field | success rate (%): 16.67 -> 33.33 | Improved |
| use Simplex ETF neural-collapse pretraining (NC-ETF) instead of none | NoMaD · Citysim Multi-Intersection | success rate (%): 48 -> 40 | Hurt |
| add ordinal-collapse geometry instead of CE-only direction labels (ablation) | ORION · Citysim Multi-Intersection | success rate (%): 16 -> 74 | Improved |
| add cross-entropy direction pretraining (ablation) | NoMaD · Citysim Multi-Intersection | success rate (%): 48 -> 16 | Hurt |
| add ordinal loss to CDNV loss (ablation) | ORION · Citysim Multi-Intersection | success rate (%): 33 -> 45 | Improved |
| add orthogonal compactness loss (ablation) | ORION · Citysim Multi-Intersection | success rate (%): 45 -> 74 | Improved |
| use two-stage instead of joint training (ablation) | ORION · Citysim Basic | success rate (%): 0 -> 94 | Improved |
| add ORION pretraining | NoMaD · Real-world Indoor | success rate (%): 40 -> 60 | Improved |
| add ORION pretraining | NoMaD · Citysim Basic | success rate (%): 88 -> 94 | Improved |
Closest papers by meaning
- Seeing through Uncertainty: Robust Task-Oriented Optimization in Visual Navigation (NeurIPS 2025 · similarity 0.88)
- DecoVLN: Decoupling Observation, Reasoning, and Correction for Vision-and-Language Navigation (CVPR 2026 · similarity 0.88)
- FocusNav: Learning Task-directed Perception via Action-Aware Future Reconstruction in Vision-and-Language Navigation (NeurIPS 2026 · similarity 0.88)
- Grounding Driving VLA via Inverse Kinematics (NeurIPS 2026 · similarity 0.86)
- NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models (AAAI 2024 · similarity 0.86)
- End-to-End (Instance)-Image Goal Navigation through Correspondence as an Emergent Phenomenon (ICLR 2024 · 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.
