YOLO11 n,s,m,l,x mAP Val500 Coco 2017 with Orange Pi 3B RK3566 NPU

This dataset contains evaluation results (in plain text format) from a custom YOLOv11 object detection model, tested on a validation set of 500 COCO-format images. The evaluation covers five YOLOv11 model variants, each representing a different trade-off between speed and accuracy: yolo11n (nano) yolo11s (small) yolo11m (medium) yolo11l (large) yolo11x (x-large) Each result file includes detailed metrics such as precision, recall, mean Average Precision (mAP), and per-class performance based on the COCO object categories. All experiments were conducted using: Ultralytics YOLO v8.3.169 Python 3.10 Torch 2.2.0 Running entirely on a CPU (ARM Cortex-A55) platform, specifically the Orange Pi 3B with RK3566 SoC. The purpose of this dataset is to benchmark the accuracy and computational performance of YOLOv11 models across different sizes in a constrained CPU-only environment. It is intended for use in academic comparisons, lightweight deployment evaluations, or model selection for edge AI applications.本数据集包含自定义YOLOv11目标检测模型的评估结果（纯文本格式），该模型在包含500张COCO格式图像的验证集上完成测试。 本次评估覆盖5种YOLOv11模型变体，每一种均代表不同的速度与精度权衡方案： yolo11n（纳米版） yolo11s（小版） yolo11m（中版） yolo11l（大版） yolo11x（超大版） 每个结果文件均包含详细指标，如精确率、召回率、平均精度均值（mean Average Precision，mAP），以及基于COCO目标类别的逐类别性能表现。 所有实验均基于以下配置： Ultralytics YOLO v8.3.169 Python 3.10 Torch 2.2.0 本次实验全程运行于ARM Cortex-A55 CPU平台，具体为搭载RK3566系统级芯片的Orange Pi 3B开发板。 本数据集旨在于受限纯CPU环境下，针对不同尺寸的YOLOv11模型开展精度与计算性能的基准测试，可用于学术对比研究、轻量化部署评估或边缘AI应用的模型选型工作。