Research map / Vision-language and multimodal / Cross-modal learning
AVION: Aerial Vision-Language Instruction from Offline Teacher to Prompt-Tuned Network
Yu Hu, Jianyang Gu, Hao Liu, Yue Cao, Jozsef Hamari, Zheng Liu, Mohsen Zardadi
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 deep prompt tuning to shallow prompt baseline (ablation) | AVION B0 · Base-to-novel (6 RS datasets) | harmonic mean: 78.88 -> 66.71 | Hurt |
| add deep prompt tuning (novel classes) (ablation) | AVION B0 · Base-to-novel (6 RS datasets) | novel accuracy: 69.52 -> 53.51 | Hurt |
| add visual alignment Limg (ablation) | AVION B1 · Base-to-novel (6 RS datasets) | harmonic mean: 66.71 -> 72.74 | Improved |
| add textual alignment with manual prototypes (ablation) | AVION B2 · Base-to-novel (6 RS datasets) | harmonic mean: 72.74 -> 78.34 | Improved |
| replace manual prototypes with LLM-generated ones (ablation) | AVION B3 · Base-to-novel (6 RS datasets) | harmonic mean: 78.34 -> 81.67 | Improved |
| add selective prototype aggregation (ablation) | AVION B4 · Base-to-novel (6 RS datasets) | harmonic mean: 81.67 -> 83.05 | Improved |
| add similarity logit alignment Llogit (ablation) | AVION B5 · Base-to-novel (6 RS datasets) | harmonic mean: 83.05 -> 86.26 | Improved |
| add 30% warm-up schedule for logit loss (ablation) | AVION B6 · Base-to-novel (6 RS datasets) | harmonic mean: 86.26 -> 87.05 | Improved |
| use domain-constrained LLM descriptions instead of manual multi-templates (ablation) | AVION · Prototype ablation (HM) | harmonic mean: 79.91 -> 81.70 | Improved |
| use selective aggregation instead of simple mean (ablation) | AVION · Prototype ablation (HM) | harmonic mean: 81.70 -> 83.05 | Improved |
| use vision-guided instead of text-only selection (ablation) | AVION · Prototype ablation (HM) | harmonic mean: 82.05 -> 83.05 | Improved |
| select random subset of LLM candidates instead of averaging all (ablation) | AVION · Prototype ablation (HM) | harmonic mean: 80.85 -> 80.31 | Hurt |
Later papers that compared with this method
1 later accepted paper reports AVION as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- LLaVAction: evaluating and training multi-modal large language models for action understanding (ICLR 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PromptKD — PromptKD: Unsupervised Prompt Distillation for Vision-Language Models (CVPR 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Remote Sensing Vision-Language Foundation Models without Annotations via Ground Remote Alignment (ICLR 2024 · similarity 0.88)
- SkySense-O: Towards Open-World Remote Sensing Interpretation with Vision-Centric Visual-Language Modeling (CVPR 2025 · similarity 0.88)
- SkyMoE: A Vision-Language Foundation Model for Enhancing Geospatial Interpretation with Mixture of Experts (AAAI 2026 · similarity 0.88)
- VRSBench: A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding (NeurIPS 2024 · similarity 0.88)
- SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensing (AAAI 2024 · similarity 0.88)
- Quality-Driven Curation of Remote Sensing Vision-Language Data via Learned Scoring Models (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.
