对比视觉语言模型与CNN在卫星地面网络频谱管理中的表现，发现二者互补而非替代。
When Does Multimodal AI Help? Diagnostic Complementarity of Vision-Language Models and CNNs for Spectrum Management in Satellite-Terrestrial Networks
- 用多粒度问答数据集诊断VLM与CNN的差异，识别任务适配性。
- CNN在空间定位上达0.552 IoU，VLM在语义推理中仅用3个示例就实现F1=0.576。
- 提出任务路由机制，融合两者优势，综合得分提升39.1%。
视觉语言模型（VLMs）在无线网络管理中的应用加速，但对其与轻量级卷积神经网络（CNN）在频谱相关任务中的优劣尚无系统理解。本文首次对非地面网络与地面网络（NTN-TN）协同系统中的频谱热图理解任务，开展VLM与CNN的诊断性比较。我们构建了SpectrumQA基准，包含108K个跨四粒度层级的视觉问答对：场景分类（L1）、区域推理（L2）、空间定位（L3）和语义推理（L4）。在三个NTN-TN场景下，基于冻结的Qwen2-VL-7B和训练好的ResNet-18的实验显示，二者呈现明确的任务依赖互补性：CNN在严重程度分类（L1）中准确率达72.9%，在空间定位（L3）中达到0.552 IoU；而VLM仅需三个上下文示例即可实现语义推理（L4）的F1=0.576，此能力在CNN中完全缺失。链式思维（CoT）提示使VLM推理提升12.6%（F1: 0.209→0.233），但对空间任务无影响，证实互补性源于架构本质而非提示限制。一个确定性任务路由器将监督任务交由CNN，推理任务交由VLM，获得复合得分0.616，较单独使用CNN提升39.1%。此外，VLM表示在跨场景迁移中表现出更强鲁棒性，在6个迁移方向中有5个性能下降更小。研究结果提供明确指导：应根据任务类型部署不同模型，而非将其视为替代品。
原文摘要 · Abstract (English)
The adoption of vision-language models (VLMs) for wireless network management is accelerating, yet no systematic understanding exists of where these large foundation models outperform lightweight convolutional neural networks (CNNs) for spectrum-related tasks. This paper presents the first diagnostic comparison of VLMs and CNNs for spectrum heatmap understanding in non-terrestrial network and terrestrial network (NTN-TN) cooperative systems. We introduce SpectrumQA, a benchmark comprising 108K visual question-answer pairs across four granularity levels: scene classification (L1), regional reasoning (L2), spatial localization (L3), and semantic reasoning (L4). Our experiments on three NTN-TN scenarios with a frozen Qwen2-VL-7B and a trained ResNet-18 reveal a clear taskdependent complementarity: CNN achieves 72.9% accuracy at severity classification (L1) and 0.552 IoU at spatial localization (L3), while VLM uniquely enables semantic reasoning (L4) with F1=0.576 using only three in-context examples-a capability fundamentally absent in CNN architectures. Chain-of-thought (CoT) prompting further improves VLM reasoning by 12.6% (F1: 0.209->0.233) while having zero effect on spatial tasks, confirming that the complementarity is rooted in architectural differences rather than prompting limitations. A deterministic task-type router that delegates supervised tasks to CNN and reasoning tasks to VLM achieves a composite score of 0.616, a 39.1% improvement over CNN alone. We further show that VLM representations exhibit stronger cross-scenario robustness, with smaller performance degradation in 5 out of 6 transfer directions. These findings provide actionable guidelines: deploy CNNs for spatial localization and VLMs for semantic spectrum reasoning, rather than treating them as substitutes.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。