确定版本

安装mmcv1.4

首先.cuda,pytorch得安装好，这里我拷贝pt1.8虚拟环境

安装mmcv1.4

安装mmdetection

参考文章


下载

cd E:\Code\mmdetection\mmdetection-2.22.0
pip install -r requirements/build.txt
conda install -c conda-forge pycocotools
pip install -v -e .

测试

遇到问题
解决mmdetection进行测试不出图片

方案1

测试代码如下

import os

os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"

from mmdet.apis import init_detector, inference_detector, show_result_pyplot
import mmcv
config_file = 'configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py'
checkpoint_file = 'checkpoints/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth'
device = 'cuda:0'
model = init_detector(config_file, checkpoint_file, device=device)
# 这一步会输出： "Use load_from_local loader"
img = 'demo/demo.jpg'
result = inference_detector(model,img)
show_result_pyplot(model,img,result)

方案2

python demo/image_demo.py demo/demo.jpg configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py checkpoints/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth