pytorch_yolov3实验记录
pytorch_yolov3实验记录
- 安装
- git项目
- 配置coco数据集
- 修改data/coco2014.data文件
- 模型测试
- 输出
- 测试结果展示
- 小结
安装
环境介绍
Ubuntu18.04
Anaconda
Python 3.7.7
GCC 7.3.0
torch 1.5
pycocotools 2.0
*必须使用numpy ==1.17.5
GPU 2070S 8G
pip install -r requirements.txt
git项目
git clone https://github.com/ultralytics/yolov3.git
配置coco数据集
cd yolov3
mkdir coco
coco路径展示
-------coco
/images
/trian2014
/val2014
/label
/train2014
/val2014
|--5k.part
|--tranvalo5k.part
cd coco
coco2014下载路径：https://pjreddie.com/media/files/train2014.zip
https://pjreddie.com/media/files/val2014.zip
这个是Darknet的镜像路径，使用迅雷下载，速度比官方的速度快
解压文件：unzip -q train2014.zip
unzip -q val2014.zip
下载注释文件：https://pjreddie.com/media/files/instances_train-val2014.zip
https://pjreddie.com/media/files/coco/5k.part
https://pjreddie.com/media/files/coco/trainvalno5k.part
https://pjreddie.com/media/files/coco/labels.tgz
tar xzf labels.tgz
unzip -q instances_train-val2014.zip
*在coco文件夹下执行
paste <(awk "{print \"$PWD\"}" <5k.part) 5k.part | tr -d '\t' > 5k.txt
paste <(awk "{print \"$PWD\"}" <trainvalno5k.part) trainvalno5k.part | tr -d '\t' > trainvalno5k.txt
修改data/coco2014.data文件
原来内容展示
classes=80
train=../coco/train2014.txt
valid=../coco/val2014.txt
names=data/coco.names
**修改路径,建议使用绝对路径
train= ../coco/trainvalno5k.txt
valid= ../coco/5k.txt
模型测试
cd yolov3
python3 test.py --data data/coco2014.data --cfg cfg/yolov3-spp.cfg --weights weights/yolov3/yolov3-spp.pt --save-json --img-size 320
输出
Namespace(augment=False, batch_size=16, cfg='cfg/yolov3.cfg', conf_thres=0.001, data='data/coco2014.data', device='', img_size=320, iou_thres=0.6, save_json=True, single_cls=False, task='test', weights='weights/yolov3/yolov3.pt')
Using CUDA device0 _CudaDeviceProperties(name='GeForce RTX 2070 SUPER', total_memory=7979MB)
Model Summary: 222 layers, 6.19491e+07 parameters, 6.19491e+07 gradients
Fusing layers...
Model Summary: 150 layers, 6.19228e+07 parameters, 6.19228e+07 gradients
Caching labels /home/gob/yolov3/coco/labels/val2014.npy (4954 found, 0 missing, 46 empty, 197 duplicate, for 5000 images): 100%|█████████████████| 5000/5000 [00:00<00:00, 26131.86it/s]
Class Images Targets P R [email protected] F1: 100%|███████████████████████████████████████████████████████████████| 313/313 [00:51<00:00, 6.05it/s]
测试结果展示
yolov3-spp测试结果
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.314
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.534
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.331
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.105
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.348
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.489
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.278
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.446
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.490
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.250
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.558
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.688
yolov3测试结果
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.295
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.521
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.306
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.111
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.331
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.455
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.265
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.430
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.474
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.236
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.545
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.661
小结
之前下载了来路不明的coco数据集，确缺少了很多图片。在这里也遇到了很多报错，但都一步步的解决了，除了基本的配置问题，路径问题之外，也对test.py和train.py文件的代码进行了修改。后期还会继续记录实验过程。。。。