文章于2020年已更新

http://m.itdecent.cn/p/2206db894b28

工具準(zhǔn)備

Darknet-YOLO:https://pjreddie.com/darknet/yolo/
labelImg:https://github.com/tzutalin/labelImg

創(chuàng)建文件夾

在darknet/scripts目錄下創(chuàng)建以下目錄

├── VOCdevkit
│   └── VOC2007
│       ├── Annotations
│       │   ├── 0a0a0b1a-7c39d841.xml
│       │   └── lena.xml
│       ├── ImageSets
│       │   ├── Layout
│       │   ├── Main
│       │   │   ├── test.txt
│       │   │   ├── train.txt
│       │   │   └── val.txt
│       │   └── Segmentation
│       ├── JPEGImages
│       │   ├── 0a0a0b1a-7c39d841.jpg
│       │   └── lena.jpg
│       └── labels
│           └── 0a0a0b1a-7c39d841.txt
└── voc_label.py

其中
JPEGImages下為訓(xùn)練測試集圖片

Annotations下為VOC格式的xml標(biāo)注
如

<annotation>
    <folder>JPEGImages</folder>
    <filename>0a0a0b1a-7c39d841.jpg</filename>
    <path>/home/dew/CV2018/yolo/darknet/scripts/VOCdevkit/VOC2007/JPEGImages/0a0a0b1a-7c39d841.jpg</path>
    <source>
        <database>Unknown</database>
    </source>
    <size>
        <width>1280</width>
        <height>720</height>
        <depth>3</depth>
    </size>
    <segmented>0</segmented>
    <object>
        <name>car</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>557</xmin>
            <ymin>275</ymin>
            <xmax>688</xmax>
            <ymax>398</ymax>
        </bndbox>
    </object>
    <object>
        <name>car</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>160</xmin>
            <ymin>297</ymin>
            <xmax>252</xmax>
            <ymax>373</ymax>
        </bndbox>
    </object>
    <object>
        <name>car</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>392</xmin>
            <ymin>298</ymin>
            <xmax>459</xmax>
            <ymax>353</ymax>
        </bndbox>
    </object>
    <object>
        <name>car</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>492</xmin>
            <ymin>304</ymin>
            <xmax>523</xmax>
            <ymax>345</ymax>
        </bndbox>
    </object>
</annotation>

Main下txt文件為對應(yīng)的測試、訓(xùn)練文件名稱
如：

0a0a0b1a-7c39d841

轉(zhuǎn)換標(biāo)注集格式

修改voc_label.py, 如只有一個(gè)class:car

sets=[('2007', 'train'), ('2007', 'val'), ('2007', 'test')]

classes = ["car"]
'''
classes = ["aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"]
'''

運(yùn)行文件script/voc_label.py

python ./voc_label.py

會(huì)在目錄下生成一系列文件并將VOC格式標(biāo)注轉(zhuǎn)為YOLO格式txt標(biāo)注（歸一化處理）見/darknet/scripts/VOCdevkit/VOC2007/labels/0a0a0b1a-7c39d841.txt

0 0.485546875 0.465972222222 0.10234375 0.170833333333
0 0.16015625 0.463888888889 0.071875 0.105555555556
0 0.331640625 0.450694444444 0.05234375 0.0763888888889
0 0.395703125 0.449305555556 0.02421875 0.0569444444444

修改cfg/voc.data

classes= 1
train  = /home/dew/Desktop/CV2018/yolo/darknet/scripts/2007_train.txt
valid  = /home/dew/Desktop/CV2018/yolo/darknet/scripts/2007_val.txt
names = data/voc.names
backup = backup

修改cfg/yolov3-voc.cfg

查找?guī)в衃convolutional]以及[yolo]標(biāo)簽處（共3處）
修改

classes  = 標(biāo)注種類數(shù)
filters=3*(classes+1+4)
ramdom=0  //顯存足夠1，不足夠0

修改data/voc.names

備份后將內(nèi)容修改為訓(xùn)練集classes名

下載預(yù)訓(xùn)練權(quán)重文件(只包含卷積層)并訓(xùn)練

wget https://pjreddie.com/media/files/darknet53.conv.74

./darknet detector train cfg/voc.data cfg/yolov3-voc.cfg darknet53.conv.74

log說明

Region xx: cfg文件中yolo-layer的索引；

Avg IOU:當(dāng)前迭代中，預(yù)測的box與標(biāo)注的box的平均交并比，越大越好，期望數(shù)值為1；

Class: 標(biāo)注物體的分類準(zhǔn)確率，越大越好，期望數(shù)值為1；

obj: 越大越好，期望數(shù)值為1；

No obj: 越小越好；

.5R: 以IOU=0.5為閾值時(shí)候的recall; recall = 檢出的正樣本/實(shí)際的正樣本

0.75R: 以IOU=0.75為閾值時(shí)候的recall;

count:正樣本數(shù)目。