Research map / Vision-language and multimodal / Robot action and navigation
AerialVLN: Vision-and-Language Navigation for UAVs
Shubo Liu, Hongsheng Zhang, Yuankai Qi, Peng Wang, Yanning Zhang, Qi Wu
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 Dataset Aggregation (DA) student-forcing to CMA (ablation) | CMA · AerialVLN-S Val Seen | SR (%): 3.0 -> 9.9 | Improved |
| add Dataset Aggregation to CMA (ablation) | CMA · AerialVLN-S Val Unseen | SR (%): 3.2 -> 4.5 | Improved |
| add look-ahead guidance (LAG) to CMA | CMA · AerialVLN-S Val Unseen | SR (%): 4.5 -> 5.1 | Improved |
| add look-ahead guidance (LAG) vs CMA-DA | CMA · AerialVLN-S Val Seen | SR (%): 9.9 -> 7.2 | Hurt |
| remove depth input from CMA (ablation) | CMA · AerialVLN Val Unseen | SR (%): 4.5 -> 3.2 | Hurt |
| remove RGB input from CMA (ablation) | CMA · AerialVLN Val Unseen | SR (%): 4.5 -> 3.0 | Hurt |
| remove language instruction from CMA (ablation) | CMA · AerialVLN Val Unseen | SR (%): 4.5 -> 2.1 | Hurt |
| remove all vision from CMA (ablation) | CMA · AerialVLN Val Unseen | SR (%): 4.5 -> 1.3 | Hurt |
| longer paths | Seq2Seq / CMA · AerialVLN | SR (%): 7.4% (short) -> 1.8% (long) | Hurt |
| use LingUNet without recurrence instead of Seq2Seq/CMA | LingUNet · AerialVLN-S unseen | SR (%): Test: 0.1 vs Seq2Seq 2.2 | Hurt |
Closest papers by meaning
- Towards Realistic UAV Vision-Language Navigation: Platform, Benchmark, and Methodology (ICLR 2025 · similarity 0.95)
- SatNav: A Scalable Benchmark for Long-Horizon UAV Vision-Language Navigation from Satellite Imagery (NeurIPS 2026 · similarity 0.91)
- Is your VLM Sky-Ready? A Comprehensive Spatial Intelligence Benchmark for UAV Navigation (CVPR 2026 · similarity 0.90)
- WorldVLN: Autoregressive World Action Model for Aerial Vision-Language Navigation (NeurIPS 2026 · similarity 0.89)
- FlySearch: Exploring how vision-language models explore (NeurIPS 2025 · similarity 0.87)
- UAV-Flow Colosseo: A Real-World Benchmark for Flying-on-a-Word UAV Imitation Learning (NeurIPS 2025 · similarity 0.87)
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.
