yolov10-benchmark's Introduction

yolov10-benchmark

This project aims to evaluate the performance of the YOLOV10 object detection model. We will measure only the inference speed here different hardware configurations.

cpu Intel(R) Xeon(R) W-2155 CPU @ 3.30GHz
cuda:0 _CudaDeviceProperties(name='Quadro P5000', major=6, minor=1, total_memory=16278MB, multi_processor_count=20)

And also this repo provide you a simple code to run yolov10

Install

Please refer to Yolov10 implementation

Usage

python main.py --video <video_path> --device <cpu/cuda>

Benchmark

I use Intel IoT Video, exactly this one.

CPU

    Report
        Model : yolov10n.pt
        Video : store-aisle-detection.mp4
                 fps : 59
                 (width, height) : (720.0,404.0)
                 Total frame : 3921
    Model inference stats :
        Device : cpu Intel(R) Xeon(R) W-2155 CPU @ 3.30GHz
        Avg predict time : 0.03s
        FPS : 28.63fps

GPU

   Report
        Model : yolov10n.pt
        Video : store-aisle-detection.mp4
                 fps : 59
                 (width, height) : (720.0,404.0)
                 Total frame : 3921
    Model inference stats :
        Device : cuda:0 _CudaDeviceProperties(name='Quadro P5000', major=6, minor=1, total_memory=16278MB, multi_processor_count=20)
        Avg predict time : 0.01s
        FPS : 103.38 fps

Citation

@misc{wang2024yolov10,
      title={YOLOv10: Real-Time End-to-End Object Detection}, 
      author={Ao Wang and Hui Chen and Lihao Liu and Kai Chen and Zijia Lin and Jungong Han and Guiguang Ding},
      year={2024},
      eprint={2405.14458},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Recommend Projects