代码拉取完成，页面将自动刷新
The YOLOv5 architecture is designed for efficient and accurate object detection tasks in real-time scenarios. It employs a single convolutional neural network to simultaneously predict bounding boxes and class probabilities for multiple objects within an image. The YOLOV5m is a medium-sized model.
# Install libGL
## CentOS
yum install -y mesa-libGL
## Ubuntu
apt install -y libgl1-mesa-dev
pip3 install tqdm
pip3 install onnx
pip3 install onnxsim
pip3 install ultralytics
pip3 install pycocotools
pip3 install cv2
pip3 install opencv-python==4.6.0.66
Pretrained model: https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5m.pt
Dataset: http://images.cocodataset.org/zips/val2017.zip to download the validation dataset.
mkdir checkpoints
git clone https://github.com/ultralytics/yolov5
# 切换到需要的版本分支
git checkout v6.1
# 有一些环境需要安装
wget https://ultralytics.com/assets/Arial.ttf
cp Arial.ttf /root/.config/Ultralytics/Arial.ttf
# 转换为onnx (具体实现可以参考 export.py 中的 export_onnx 函数)
python3 export.py --weights yolov5m.pt --include onnx --opset 11 --batch-size 32
mv yolov5m.onnx /Path/to/checkpoints
export PROJ_DIR=/Path/to/yolov5m/ixrt
export DATASETS_DIR=/Path/to/coco2017/
export CHECKPOINTS_DIR=./checkpoints
export COCO_GT=${DATASETS_DIR}/annotations/instances_val2017.json
export EVAL_DIR=${DATASETS_DIR}/val2017
export RUN_DIR=/Path/to/yolov5m/ixrt
export CONFIG_DIR=config/YOLOV5M_CONFIG
# Accuracy
bash scripts/infer_yolov5m_fp16_accuracy.sh
# Performance
bash scripts/infer_yolov5m_fp16_performance.sh
# Accuracy
bash scripts/infer_yolov5m_int8_accuracy.sh
# Performance
bash scripts/infer_yolov5m_int8_performance.sh
| Model | BatchSize | Precision | FPS | MAP@0.5 | MAP@0.5:0.95 |
|---|---|---|---|---|---|
| YOLOv5m | 32 | FP16 | 680.93 | 0.637 | 0.447 |
| YOLOv5m | 32 | INT8 | 1328.50 | 0.627 | 0.425 |
此处可能存在不合适展示的内容，页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容，可点击提交进行申诉，我们将尽快为您处理。
马建仓 AI 助手
