Research map / Probabilistic and scientific modeling / Molecules and proteins
Mitigating Error Accumulation in Continuous Navigation via Memory-Augmented Kalman Filtering
Yin Tang, Jiawei Ma, Jinrui Zhang, Alex Jinpeng Wang, Deyu Zhang
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 |
|---|---|---|---|
| use learnable Kalman gain instead of fixed 0.5 (ablation) | NeuroKalman · L1 Test-Seen Full | SR: 24.12 -> 25.86 | Improved |
| use learnable Kalman gain instead of fixed 0.9 (ablation) | NeuroKalman · L1 Test-Seen Full | SR: 18.05 -> 25.86 | Improved |
| use learnable Kalman gain instead of fixed 0.1 (ablation) | NeuroKalman · L1 Test-Seen Full | SR: 0.00 -> 25.86 | Improved |
| memory length 10 instead of 5 (ablation) | NeuroKalman · L1 Test-Seen Full | NE: 84.39 -> 71.56 | Improved |
| memory length 10 instead of 15 (ablation) | NeuroKalman · L1 Test-Seen Full | NE: 77.17 -> 71.56 | Improved |
| memory insertion threshold 0.5 instead of 0.3 (ablation) | NeuroKalman · L1 Test-Seen Full | SR: 20.50 -> 25.86 | Improved |
| threshold above 0.5 (0.7, 0.9) (ablation) | NeuroKalman · L1 Test-Seen Full | SR: 25.86 -> 24.55 / 24.18 | No clear effect |
| add latent Kalman correction with memory vs post-hoc constant-velocity Kalman filter on outputs (ablation) | TravelUAV · L1 Test-Seen Full | NE: 96.67 -> 71.56 | Improved |
| add post-hoc Kalman filter on waypoints (ablation) | TravelUAV · L1 Test-Seen Full | OSR: 44.26 -> 33.65 | Hurt |
| add NeuroKalman memory-augmented Kalman correction | TravelUAV-FT · L1 Test-Seen Full, 10% fine-tune | SR: 17.56 -> 25.86 | Improved |
| add NeuroKalman correction, no full-data pretraining | TravelUAV · L1 Test-Seen, 10% data | SR: 13.19 -> 22.70 | Improved |
| add NeuroKalman correction | TravelUAV · L1 Test-Unseen-Map Full | SR: 4.18 -> 8.34 | Improved |
Closest papers by meaning
- WorldVLN: Autoregressive World Action Model for Aerial Vision-Language Navigation (NeurIPS 2026 · similarity 0.86)
- SatNav: A Scalable Benchmark for Long-Horizon UAV Vision-Language Navigation from Satellite Imagery (NeurIPS 2026 · similarity 0.85)
- Towards Realistic UAV Vision-Language Navigation: Platform, Benchmark, and Methodology (ICLR 2025 · similarity 0.85)
- Parse, Search, and Confirmation: Training-Free Aerial Vision-and-Dialog Navigation with Chain-of-Thought Reasoning and Structured Spatial Memory (CVPR 2026 · similarity 0.84)
- C-NAV: Towards Self-Evolving Continual Object Navigation in Open World (NeurIPS 2025 · similarity 0.84)
- Active Test-time Vision-Language Navigation (NeurIPS 2025 · similarity 0.84)
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.
