caffe-ssd 训练自己的模型

需要修改的地方是红线部分，根据分类数量进行修改，原版的开源是21个分类（包括background），所以是63（213）,126（213*2）,如果class_num为5，则为15，30

1、首先准备好自己的训练数据，包括，Annotations和JPEGImages，ImageSets通过如下python脚本和前面两个文件生成：

# _*_ coding:UTF-8 _*_
#开发作者   :   ZhangRong 
#开发时间   :   2020/4/21  10:18
#文件名称   :   process.py
#开发工具   :   PyCharm
#Description:

import os
import random
trainval_percent = 0.66
train_percent = 0.8
xmlfilepath = '/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/MyDataSet/Annotations'
txtsavepath = '/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdeckit/MyDataSet/ImageSets/Main'
total_xml = os.listdir(xmlfilepath)
num=len(total_xml)
list=range(num)
tv=int(num*trainval_percent)
tr=int(tv*train_percent)
trainval= random.sample(list,tv)
train=random.sample(trainval,tr)
ftrainval = open('/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/MyDataSet/ImageSets/Main/trainval.txt', 'w')
ftest = open('/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/MyDataSet/ImageSets/Main/test.txt', 'w')
ftrain = open('/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/MyDataSet/ImageSets/Main/train.txt', 'w')
fval = open('/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/MyDataSet/ImageSets/Main/val.txt', 'w')
for i in list:
	name=total_xml[i][:-4]+'\n'
	if i in trainval:
		ftrainval.write(name)
		if i in train:
			ftrain.write(name)
		else:
			fval.write(name)
	else:
		ftest.write(name)
ftrainval.close()
ftrain.close()
fval.close()
ftest .close()

2、生成lmdb，如下两个脚本
create_list.sh

#!/usr/local/bin/bash
root_dir=/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit/  # 修改成自己的路径
sub_dir=ImageSets/Main
bash_dir="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
for dataset in trainval test
do
  dst_file=$bash_dir/$dataset.txt
  if [ -f $dst_file ]
  then
    rm -f $dst_file
  fi
  for name in MyDataSet  # 也需要修改
  do
    if [[ $dataset == "test" && $name == "VOC2012" ]]
    then
      continue
    fi
    echo "Create list for $name $dataset..."
    dataset_file=$root_dir/$name/$sub_dir/$dataset.txt
    img_file=$bash_dir/$dataset"_img.txt"
    cp $dataset_file $img_file
    sed -i "s/^/$name\/JPEGImages\//g" $img_file
    sed -i "s/$/.jpg/g" $img_file
 
    label_file=$bash_dir/$dataset"_label.txt"
    cp $dataset_file $label_file
    sed -i "s/^/$name\/Annotations\//g" $label_file
    sed -i "s/$/.xml/g" $label_file
    paste -d' ' $img_file $label_file >> $dst_file
    rm -f $label_file
    rm -f $img_file
  done
  # Generate image name and size infomation.
  if [ $dataset == "test" ]
  then
    $bash_dir/../../build/tools/get_image_size $root_dir $dst_file $bash_dir/$dataset"_name_size.txt"
  fi
  # Shuffle trainval file.
  if [ $dataset == "trainval" ]
  then
    rand_file=$dst_file.random
    cat $dst_file | perl -MList::Util=shuffle -e 'print shuffle();' > $rand_file
    mv $rand_file $dst_file
  fi
done

crea_data.sh

cur_dir=$(cd $( dirname ${BASH_SOURCE[0]} ) && pwd )
root_dir="/data/zhangrong/caffe/caffe-ssd/caffe"

cd $root_dir

redo=1
data_root_dir="/data/zhangrong/caffe/caffe-ssd/caffe/data/VOCdevkit"
dataset_name="MyDataSet"
mapfile="$root_dir/data/$dataset_name/labelmap_voc.prototxt"
anno_type="detection"
db="lmdb"
min_dim=0
max_dim=0
width=300 #resize大小
height=300

extra_cmd="--encode-type=jpg --encoded"
if [ $redo ]
then
  extra_cmd="$extra_cmd --redo"
fi
for subset in test trainval
do
  python $root_dir/scripts/create_annoset.py --anno-type=$anno_type --label-map-file=$mapfile --min-dim=$min_dim --max-dim=$max_dim --resize-width=$width --resize-height=$height --check-label $extra_cmd $data_root_dir $root_dir/data/$dataset_name/$subset.txt $data_root_dir/$dataset_name/$db/$dataset_name"_"$subset"_"$db examples/$dataset_name
done

3、训练脚本
train.sh

#!/usr/local/bin/bash
if ! test -f MobileNetSSD_train.prototxt ;then
	echo "error: train.prototxt does not exist."
  exit 1
fi

echo $(ls -t snapshot/voc/*.caffemodel | head -n 1)

#set your caffe path, such as '/home/work/caffe_ssd/build/tools/caffe'
/data/zhangrong/caffe/caffe-ssd/caffe/build/tools/caffe train --solver="solver.prototxt" \
--weights=$(ls -t backup/*.caffemodel | head -n 1) \
--gpu 0 2>&1 | tee backup/MobileNetSSD_voc.log