Research map / Vision-language and multimodal / Multimodal LLMs
NeSy-Route: A Neural-Symbolic Benchmark for Constrained Route Planning in Remote Sensing
Ming Yang, Zhi Zhou, Shi-Yu Tian, Kun-Yang Yu, Lan-Zhe Guo, Yu-Feng Li
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 visual input (Task 1 text only to Task 2 image+text) | MLLMs · NeSy-Route | TM: Gemini-3-Pro 98.34 -> 27.50; GPT-5.1 77.02 -> 35.31 | Hurt |
| Increase scene complexity Easy to Hard | GPT-5.1 · NeSy-Route Task 3 | AR: 25.90 -> 12.10 | Hurt |
| Increase scene complexity Easy to Hard | GPT-5.1 · NeSy-Route Task 2 | TM: 66.37 -> 16.72 | Hurt |
| Use Qwen3-VL-235B-A22B MoE instead of GPT-5.1 | Task 2 alignment · NeSy-Route Task 2 | TM: 35.31 -> 52.79 | Improved |
| Scale Qwen3-VL from 8B to 235B-A22B | Qwen3-VL · NeSy-Route Task 1 | FM: 27.67 -> 52.98 | Improved |
| Scale Qwen3-VL from 30B-A3B to 32B dense | Qwen3-VL · NeSy-Route Task 1 | PR: 0.752 -> -0.616 | Hurt |
| Compare Qwen3-VL-32B against Gemini-3-Pro on planning | Task 3 · NeSy-Route Task 3 | AR vs CR: AR 32.70 vs 18.70; CR 12.87 vs 1.24 | Mixed |
| Strong Task 1 performance as prerequisite for Task 2 | LLaVA-OneVision · NeSy-Route Task 2 | TM: Task 1 TM 26.68; Task 2 avg TM 14.52 | Hurt |
| Strong perception/reasoning vs planning | Gemini-3-Pro · NeSy-Route | FM/AR: Task 1 FM 92.24 vs Task 3 AR 18.70 | Mixed |
Closest papers by meaning
- Towards Faithful Reasoning in Remote Sensing: A Perceptually-Grounded GeoSpatial Chain-of-Thought for Vision-Language Models (ICLR 2026 · similarity 0.89)
- CHOICE: Benchmarking the Remote Sensing Capabilities of Large Vision-Language Models (NeurIPS 2025 · similarity 0.88)
- VLRS-Bench: A Vision-Language Reasoning Benchmark for Remote Sensing (NeurIPS 2026 · similarity 0.87)
- GeoCoT: Towards Reliable Remote Sensing Reasoning with Manifold Perspective (CVPR 2026 · similarity 0.86)
- RemoteReasoner: Towards Unifying Geospatial Reasoning Workflow (AAAI 2026 · similarity 0.86)
- SkyMoE: A Vision-Language Foundation Model for Enhancing Geospatial Interpretation with Mixture of Experts (AAAI 2026 · 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.
