yolo-tiny的训练

yolov2-tiny和yolov3-tiny由于速度快，占内存少等优点，被广泛使用

yolov2-tiny的训练

登录服务器

ssh zibojia@...org

使用yolov2-tiny.cfg

打开，进行更改

# Testing
# batch=1
# subdivisions=1
# Training
batch=64
subdivisions=2
width=416
height=416
channels=3
momentum=0.9
decay=0.0005
angle=0
saturation = 1.5
exposure = 1.5
hue=.1

创建facial_expression.data

classes= 1
train  = /home/users/zibojia/fddb_train.txt
valid  = /home/users/zibojia/fddb_test.txt
names = cfg/face.names
backup = /home/users/zibojia

创建face.names

happiness
anger
fear
sad
disgust
surprise
neutral

更改[convolutional]

设置filters为18

filters=18

更改[r