arXiv · cs.AI · Zhengshu Zhou, Shaoyu Wang, Ziyi Geng, Lu Tao, Qian Long · 8 小时前模型评分32
HydroReasoner：面向云边协同监测的知识引导积水水位推理框架
Knowledge-Guided Floodwater Level Reasoning for Cloud-Edge Collaborative Monitoring
导读
研究者提出积水水位推理框架 HydroReasoner，基于 Qwen2.5-VL 并用 Group Relative Policy Optimization 训练，借助积水领域知识图谱 HydroKG 的路径奖励监督推理过程。
分类置信度
中 · 待人工复核
- 存在未解决的相邻类别：data-resources、embodied-ai，需要人工复核。
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
应用领域：交通与物流
事件类型：研究发布
证据摘录
- “To address this, we reformulate the task as an explicit visual reasoning problem and propose HydroReasoner, a floodwater level reasoning framework designed for cloud-edge collaboration.”
- “For edge deployment, we apply 4-bit quantization to the model, cutting GPU memory usage by 57% and inference latency by 43% while out-of-domain accuracy drops by only 0.2 percentage points.”
来源：arXiv · cs.AI · arxiv.org