探索基于生成式AI的图像修复与农作物检测中的模型量化
Exploring Model Quantization in GenAI-based Image Inpainting and Detection of Arable Plants
AI总结：
本研究提出基于Stable Diffusion的数据增强框架，通过FP16和INT8量化策略平衡YOLO11和RT-DETR的推理速度与精度，并在Jetson Orin Nano上验证了其在资源受限环境下提升杂草检测性能的可行性。
摘要：
基于深度学习的杂草控制系统常受限于训练数据多样性不足和机载计算能力受限，影响其实际性能。为克服这些挑战，我们提出一种利用基于Stable Diffusion的修复技术，以10%增量逐步增强训练数据——最高增加200%，从而提升样本的数量与多样性。我们的方法在两个最先进的目标检测模型YOLO11(l)和RT-DETR(l)上进行评估，使用mAP50指标衡量检测性能。我们为生成式修复和检测模型探索量化策略（FP16和INT8），以平衡推理速度与精度。在Jetson Orin Nano上部署下游模型，验证了本框架在资源受限环境中的实际可行性，最终提升了智能杂草管理系统的检测精度与计算效率。
英文摘要：
Deep learning-based weed control systems often suffer from limited training data diversity and constrained on-board computation, impacting their real-world performance. To overcome these challenges, we propose a framework that leverages Stable Diffusion-based inpainting to augment training data progressively in 10% increments -- up to an additional 200%, thus enhancing both the volume and diversity of samples. Our approach is evaluated on two state-of-the-art object detection models, YOLO11(l) and RT-DETR(l), using the mAP50 metric to assess detection performance. We explore quantization strategies (FP16 and INT8) for both the generative inpainting and detection models to strike a balance between inference speed and accuracy. Deployment of the downstream models on the Jetson Orin Nano demonstrates the practical viability of our framework in resource-constrained environments, ultimately improving detection accuracy and computational efficiency in intelligent weed management systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
