SSD demo中详细介绍了如何在VOC数据集上使用SSD进行物体检测的训练和验证。本文介绍如何使用SSD实现对自己数据集的训练和验证过程，内容包括：
1 数据集的标注2 数据集的转换3 使用SSD如何训练4 使用SSD如何测试 1 数据集的标注 数据的标注使用BBox-Label-Tool工具，该工具使用python实现，使用简单方便。该工具生成的标签格式是:object_numberx1min y1min x1max y1maxx2min y2min x2max y2max...2 数据集的转换 caffe训练使用LMDB格式的数据，ssd框架中提供了voc数据格式转换成LMDB格式的脚本。所以实践中先将BBox-Label-Tool标注的数据转换成voc数据格式，然后再转换成LMDB格式。 2.1 voc数据格式
(1)Annotations中保存的是xml格式的label信息
<?xml version="1.0" ?>
<annotation>
<folder>VOC2007</folder>
<filename>1.jpg</filename>
<source>
<database>My Database</database>
<annotation>VOC2007</annotation>
<image>flickr</image>
<flickrid>NULL</flickrid>
</source>
<owner>
<flickrid>NULL</flickrid>
<name>idaneel</name>
</owner>
<size>
<width>320</width>
<height>240</height>
<depth>3</depth>
</size>
<segmented>0</segmented>
<object>
<name>door</name>
<pose>Unspecified</pose>
<truncated>0</truncated>
<difficult>0</difficult>
<bndbox>
<xmin>109</xmin>
<ymin>3</ymin>
<xmax>199</xmax>
<ymax>204</ymax>
</bndbox>
</object>
</annotation>
VOC XML内容信息
(2)ImageSet目录下的Main目录里存放的是用于表示训练的图片集和测试的图片集
(3)JPEGImages目录下存放所有图片集
(4)label目录下保存的是BBox-Label-Tool工具标注好的bounding box坐标文件，该目录下的文件就是待转换的label标签文件。
2.2 Label转换成VOC数据格式
BBox-Label-Tool工具标注好的bounding box坐标文件转换成VOC数据格式的形式.具体的转换过程包括了两个步骤：（1）将BBox-Label-Tool下的txt格式保存的bounding box信息转换成VOC数据格式下以xml方式表示；（2）生成用于训练的数据集和用于测试的数据集。用python实现了上述两个步骤的换转。createXml.py 完成txt到xml的转换； 执行脚本./createXml.py %classname%
#!/usr/bin/env python
import os
import sys
import cv2
from itertools import islice
from xml.dom.minidom import Document
labels=‘label‘
imgpath=‘JPEGImages/‘
xmlpath_new=‘Annotations/‘
foldername=‘VOC2007‘
try:
labelName = sys.argv[1]
except:
print ‘Please input class name‘
print ‘./createXml dog‘
os._exit(0)
def insertObject(doc, datas):
obj = doc.createElement(‘object‘)
name = doc.createElement(‘name‘)
name.appendChild(doc.createTextNode(labelName))
obj.appendChild(name)
pose = doc.createElement(‘pose‘)
pose.appendChild(doc.createTextNode(‘Unspecified‘))
obj.appendChild(pose)
truncated = doc.createElement(‘truncated‘)
truncated.appendChild(doc.createTextNode(str(0)))
obj.appendChild(truncated)
difficult = doc.createElement(‘difficult‘)
difficult.appendChild(doc.createTextNode(str(0)))
obj.appendChild(difficult)
bndbox = doc.createElement(‘bndbox‘)
xmin = doc.createElement(‘xmin‘)
xmin.appendChild(doc.createTextNode(str(datas[0])))
bndbox.appendChild(xmin)
ymin = doc.createElement(‘ymin‘)
ymin.appendChild(doc.createTextNode(str(datas[1])))
bndbox.appendChild(ymin)
xmax = doc.createElement(‘xmax‘)
xmax.appendChild(doc.createTextNode(str(datas[2])))
bndbox.appendChild(xmax)
ymax = doc.createElement(‘ymax‘)
ymax.appendChild(doc.createTextNode(str(datas[3])[0:-1]))
bndbox.appendChild(ymax)
obj.appendChild(bndbox)
return obj
def create():
for walk in os.walk(labels):
for each in walk[2]:
fidin=open(walk[0] + ‘/‘+ each,‘r‘)
objIndex = 0
for data in islice(fidin, 1, None):
objIndex += 1
data=data.strip(‘\n‘)
datas = data.split(‘ ‘)
pictureName = each.replace(‘.txt‘, ‘.jpg‘)
imageFile = imgpath + pictureName
img = cv2.imread(imageFile)
imgSize = img.shape
if 1 == objIndex:
xmlName = each.replace(‘.txt‘, ‘.xml‘)
f = open(xmlpath_new + xmlName, "w")
doc = Document()
annotation = doc.createElement(‘annotation‘)
doc.appendChild(annotation)
folder = doc.createElement(‘folder‘)
folder.appendChild(doc.createTextNode(foldername))
annotation.appendChild(folder)
filename = doc.createElement(‘filename‘)
filename.appendChild(doc.createTextNode(pictureName))
annotation.appendChild(filename)
source = doc.createElement(‘source‘)
database = doc.createElement(‘database‘)
database.appendChild(doc.createTextNode(‘My Database‘))
source.appendChild(database)
source_annotation = doc.createElement(‘annotation‘)
source_annotation.appendChild(doc.createTextNode(foldername))
source.appendChild(source_annotation)
image = doc.createElement(‘image‘)
image.appendChild(doc.createTextNode(‘flickr‘))
source.appendChild(image)
flickrid = doc.createElement(‘flickrid‘)
flickrid.appendChild(doc.createTextNode(‘NULL‘))
source.appendChild(flickrid)
annotation.appendChild(source)
owner = doc.createElement(‘owner‘)
flickrid = doc.createElement(‘flickrid‘)
flickrid.appendChild(doc.createTextNode(‘NULL‘))
owner.appendChild(flickrid)
name = doc.createElement(‘name‘)
name.appendChild(doc.createTextNode(‘idaneel‘))
owner.appendChild(name)
annotation.appendChild(owner)
size = doc.createElement(‘size‘)
width = doc.createElement(‘width‘)
width.appendChild(doc.createTextNode(str(imgSize[1])))
size.appendChild(width)
height = doc.createElement(‘height‘)
height.appendChild(doc.createTextNode(str(imgSize[0])))
size.appendChild(height)
depth = doc.createElement(‘depth‘)
depth.appendChild(doc.createTextNode(str(imgSize[2])))
size.appendChild(depth)
annotation.appendChild(size)
segmented = doc.createElement(‘segmented‘)
segmented.appendChild(doc.createTextNode(str(0)))
annotation.appendChild(segmented)
annotation.appendChild(insertObject(doc, datas))
else:
annotation.appendChild(insertObject(doc, datas))
try:
f.write(doc.toprettyxml(indent = ‘ ‘))
f.close()
fidin.close()
except:
pass
if __name__ == ‘__main__‘:
create()
createXml.py
createTest.py 生成训练集和测试集标识文件； 执行脚本
./createTest.py %startID% %endID% %testNumber%
#!/usr/bin/env python
import os
import sys
import random
try:
start = int(sys.argv[1])
end = int(sys.argv[2])
test = int(sys.argv[3])
allNum = end-start+1
except:
print ‘Please input picture range‘
print ‘./createTest.py 1 1500 500‘
os._exit(0)
b_list = range(start,end)
blist_webId = random.sample(b_list, test)
blist_webId = sorted(blist_webId)
allFile = []
testFile = open(‘ImageSets/Main/test.txt‘, ‘w‘)
trainFile = open(‘ImageSets/Main/trainval.txt‘, ‘w‘)
for i in range(allNum):
allFile.append(i+1)
for test in blist_webId:
allFile.remove(test)
testFile.write(str(test) + ‘\n‘)
for train in allFile:
trainFile.write(str(train) + ‘\n‘)
testFile.close()
trainFile.close()
createTest.py
说明： 由于BBox-Label-Tool实现相对简单，该工具每次只能对一个类别进行打标签，所以转换脚本
每一次也是对一个类别进行数据的转换，这个问题后续需要优化改进。
2.3 VOC数据转换成LMDB数据
SSD提供了VOC数据到LMDB数据的转换脚本 data/VOC0712/create_list.sh 和 ./data/VOC0712/create_data.sh，这两个脚本是完全针对VOC0712目录下的数据进行的转换。 实现中为了不破坏VOC0712目录下的数据内容，针对我们自己的数据集，修改了上面这两个脚本，将脚本中涉及到VOC0712的信息替换成我们自己的目录信息。在处理我们的数据集时，将VOC0712替换成indoor。具体的步骤如下： (1) 在 $HOME/data/VOCdevkit目录下创建indoor目录，该目录中存放自己转换完成的VOC数据集； (2) $CAFFE_ROOT/examples目录下创建indoor目录； (3) $CAFFE_ROOT/data目录下创建indoor目录，同时将data/VOC0712下的create_list.sh,create_data.sh,labelmap_voc.prototxt这三个文件copy到indoor目录下，分别重命名为create_list_indoor.sh,create_data_indoor.sh, labelmap_indoor.prototxt (4)对上面新生成的两个create文件进行修改，主要修改是将VOC0712相关的信息替换成indoor 修改后的这两个文件分别为：
#!/bin/bash
root_dir=$HOME/data/VOCdevkit/
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 indoor
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(<STDIN>);‘ > $rand_file
mv $rand_file $dst_file
fi
done
create_list_indoor.sh
cur_dir=$(cd $( dirname ${BASH_SOURCE[0]} ) && pwd )
root_dir=$cur_dir/../..
cd $root_dir
redo=1
data_root_dir="$HOME/data/VOCdevkit"
dataset_name="indoor"
mapfile="$root_dir/data/$dataset_name/labelmap_indoor.prototxt"
anno_type="detection"
db="lmdb"
min_dim=0
max_dim=0
width=0
height=0
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/$dataset_name
done
create_data_indoor.sh
(5)修改labelmap_indoor.prototxt，将该文件中的类别修改成和自己的数据集相匹配，注意需要保留一个label 0 , background类别
item {
name: "none_of_the_above"
label: 0
display_name: "background"
}
item {
name: "door"
label: 1
display_name: "door"
}
labelmap_indoor.prototxt
完成上面步骤的修改后，可以开始LMDB数据数据的制作，在$CAFFE_ROOT目录下分别运行：
./data/indoor/create_list_indoor.sh
./data/indoor/create_data_indoor.sh
命令执行完毕后，可以在$CAFFE_ROOT/indoor目录下查看转换完成的LMDB数据数据。
3 使用SSD进行自己数据集的训练
训练时使用ssd demo中提供的预训练好的VGGnet model : VGG_ILSVRC_16_layers_fc_reduced.caffemodel将该模型保存到$CAFFE_ROOT/models/VGGNet下。将ssd_pascal.py copy一份 ssd_pascal_indoor.py文件， 根据自己的数据集修改ssd_pascal_indoor.py主要修改点： （1）train_data和test_data修改成指向自己的数据集LMDB train_data = "examples/indoor/indoor_trainval_lmdb" test_data = "examples/indoor/indoor_test_lmdb"（2） num_test_image该变量修改成自己数据集中测试数据的数量（3）num_classes 该变量修改成自己数据集中 标签类别数量数 + 1
针对我的数据集，ssd_pascal_indoor.py的内容为：
from __future__ import print_function
import caffe
from caffe.model_libs import *
from google.protobuf import text_format
import math
import os
import shutil
import stat
import subprocess
import sys
# Add extra layers on top of a "base" network (e.g. VGGNet or Inception).
def AddExtraLayers(net, use_batchnorm=True):
use_relu = True
# Add additional convolutional layers.
from_layer = net.keys()[-1]
# TODO(weiliu89): Construct the name using the last layer to avoid duplication.
out_layer = "conv6_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 1, 0, 1)
from_layer = out_layer
out_layer = "conv6_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 512, 3, 1, 2)
for i in xrange(7, 9):
from_layer = out_layer
out_layer = "conv{}_1".format(i)
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1)
from_layer = out_layer
out_layer = "conv{}_2".format(i)
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 1, 2)
# Add global pooling layer.
name = net.keys()[-1]
net.pool6 = L.Pooling(net[name], pool=P.Pooling.AVE, global_pooling=True)
return net
### Modify the following parameters accordingly ###
# The directory which contains the caffe code.
# We assume you are running the script at the CAFFE_ROOT.
caffe_root = os.getcwd()
# Set true if you want to start training right after generating all files.
run_soon = True
# Set true if you want to load from most recently saved snapshot.
# Otherwise, we will load from the pretrain_model defined below.
resume_training = True
# If true, Remove old model files.
remove_old_models = False
# The database file for training data. Created by data/VOC0712/create_data.sh
train_data = "examples/indoor/indoor_trainval_lmdb"
# The database file for testing data. Created by data/VOC0712/create_data.sh
test_data = "examples/indoor/indoor_test_lmdb"
# Specify the batch sampler.
resize_width = 300
resize_height = 300
resize = "{}x{}".format(resize_width, resize_height)
batch_sampler = [
{
‘sampler‘: {
},
‘max_trials‘: 1,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘min_jaccard_overlap‘: 0.1,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘min_jaccard_overlap‘: 0.3,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘min_jaccard_overlap‘: 0.5,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘min_jaccard_overlap‘: 0.7,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘min_jaccard_overlap‘: 0.9,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
{
‘sampler‘: {
‘min_scale‘: 0.3,
‘max_scale‘: 1.0,
‘min_aspect_ratio‘: 0.5,
‘max_aspect_ratio‘: 2.0,
},
‘sample_constraint‘: {
‘max_jaccard_overlap‘: 1.0,
},
‘max_trials‘: 50,
‘max_sample‘: 1,
},
]
train_transform_param = {
‘mirror‘: True,
‘mean_value‘: [104, 117, 123],
‘resize_param‘: {
‘prob‘: 1,
‘resize_mode‘: P.Resize.WARP,
‘height‘: resize_height,
‘width‘: resize_width,
‘interp_mode‘: [
P.Resize.LINEAR,
P.Resize.AREA,
P.Resize.NEAREST,
P.Resize.CUBIC,
P.Resize.LANCZOS4,
],
},
‘emit_constraint‘: {
‘emit_type‘: caffe_pb2.EmitConstraint.CENTER,
}
}
test_transform_param = {
‘mean_value‘: [104, 117, 123],
‘resize_param‘: {
‘prob‘: 1,
‘resize_mode‘: P.Resize.WARP,
‘height‘: resize_height,
‘width‘: resize_width,
‘interp_mode‘: [P.Resize.LINEAR],
},
}
# If true, use batch norm for all newly added layers.
# Currently only the non batch norm version has been tested.
use_batchnorm = False
# Use different initial learning rate.
if use_batchnorm:
base_lr = 0.0004
else:
# A learning rate for batch_size = 1, num_gpus = 1.
base_lr = 0.00004
# Modify the job name if you want.
job_name = "SSD_{}".format(resize)
# The name of the model. Modify it if you want.
model_name = "VGG_VOC0712_{}".format(job_name)
# Directory which stores the model .prototxt file.
save_dir = "models/VGGNet/VOC0712/{}".format(job_name)
# Directory which stores the snapshot of models.
snapshot_dir = "models/VGGNet/VOC0712/{}".format(job_name)
# Directory which stores the job script and log file.
job_dir = "jobs/VGGNet/VOC0712/{}".format(job_name)
# Directory which stores the detection results.
output_result_dir = "{}/data/VOCdevkit/results/VOC2007/{}/Main".format(os.environ[‘HOME‘], job_name)
# model definition files.
train_net_file = "{}/train.prototxt".format(save_dir)
test_net_file = "{}/test.prototxt".format(save_dir)
deploy_net_file = "{}/deploy.prototxt".format(save_dir)
solver_file = "{}/solver.prototxt".format(save_dir)
# snapshot prefix.
snapshot_prefix = "{}/{}".format(snapshot_dir, model_name)
# job script path.
job_file = "{}/{}.sh".format(job_dir, model_name)
# Stores the test image names and sizes. Created by data/VOC0712/create_list.sh
name_size_file = "data/indoor/test_name_size.txt"
# The pretrained model. We use the Fully convolutional reduced (atrous) VGGNet.
pretrain_model = "models/VGGNet/VGG_ILSVRC_16_layers_fc_reduced.caffemodel"
# Stores LabelMapItem.
label_map_file = "data/indoor/labelmap_indoor.prototxt"
# MultiBoxLoss parameters.
num_classes = 2
share_location = True
background_label_id=0
train_on_diff_gt = True
normalization_mode = P.Loss.VALID
code_type = P.PriorBox.CENTER_SIZE
neg_pos_ratio = 3.
loc_weight = (neg_pos_ratio + 1.) / 4.
multibox_loss_param = {
‘loc_loss_type‘: P.MultiBoxLoss.SMOOTH_L1,
‘conf_loss_type‘: P.MultiBoxLoss.SOFTMAX,
‘loc_weight‘: loc_weight,
‘num_classes‘: num_classes,
‘share_location‘: share_location,
‘match_type‘: P.MultiBoxLoss.PER_PREDICTION,
‘overlap_threshold‘: 0.5,
‘use_prior_for_matching‘: True,
‘background_label_id‘: background_label_id,
‘use_difficult_gt‘: train_on_diff_gt,
‘do_neg_mining‘: True,
‘neg_pos_ratio‘: neg_pos_ratio,
‘neg_overlap‘: 0.5,
‘code_type‘: code_type,
}
loss_param = {
‘normalization‘: normalization_mode,
}
# parameters for generating priors.
# minimum dimension of input image
min_dim = 300
# conv4_3 ==> 38 x 38
# fc7 ==> 19 x 19
# conv6_2 ==> 10 x 10
# conv7_2 ==> 5 x 5
# conv8_2 ==> 3 x 3
# pool6 ==> 1 x 1
mbox_source_layers = [‘conv4_3‘, ‘fc7‘, ‘conv6_2‘, ‘conv7_2‘, ‘conv8_2‘, ‘pool6‘]
# in percent %
min_ratio = 20
max_ratio = 95
step = int(math.floor((max_ratio - min_ratio) / (len(mbox_source_layers) - 2)))
min_sizes = []
max_sizes = []
for ratio in xrange(min_ratio, max_ratio + 1, step):
min_sizes.append(min_dim * ratio / 100.)
max_sizes.append(min_dim * (ratio + step) / 100.)
min_sizes = [min_dim * 10 / 100.] + min_sizes
max_sizes = [[]] + max_sizes
aspect_ratios = [[2], [2, 3], [2, 3], [2, 3], [2, 3], [2, 3]]
# L2 normalize conv4_3.
normalizations = [20, -1, -1, -1, -1, -1]
# variance used to encode/decode prior bboxes.
if code_type == P.PriorBox.CENTER_SIZE:
prior_variance = [0.1, 0.1, 0.2, 0.2]
else:
prior_variance = [0.1]
flip = True
clip = True
# Solver parameters.
# Defining which GPUs to use.
gpus = "0"
gpulist = gpus.split(",")
num_gpus = len(gpulist)
# Divide the mini-batch to different GPUs.
batch_size = 4
accum_batch_size = 32
iter_size = accum_batch_size / batch_size
solver_mode = P.Solver.CPU
device_id = 0
batch_size_per_device = batch_size
if num_gpus > 0:
batch_size_per_device = int(math.ceil(float(batch_size) / num_gpus))
iter_size = int(math.ceil(float(accum_batch_size) / (batch_size_per_device * num_gpus)))
solver_mode = P.Solver.GPU
device_id = int(gpulist[0])
if normalization_mode == P.Loss.NONE:
base_lr /= batch_size_per_device
elif normalization_mode == P.Loss.VALID:
base_lr *= 25. / loc_weight
elif normalization_mode == P.Loss.FULL:
# Roughly there are 2000 prior bboxes per image.
# TODO(weiliu89): Estimate the exact # of priors.
base_lr *= 2000.
# Which layers to freeze (no backward) during training.
freeze_layers = [‘conv1_1‘, ‘conv1_2‘, ‘conv2_1‘, ‘conv2_2‘]
# Evaluate on whole test set.
num_test_image = 800
test_batch_size = 1
test_iter = num_test_image / test_batch_size
solver_param = {
# Train parameters
‘base_lr‘: base_lr,
‘weight_decay‘: 0.0005,
‘lr_policy‘: "step",
‘stepsize‘: 40000,
‘gamma‘: 0.1,
‘momentum‘: 0.9,
‘iter_size‘: iter_size,
‘max_iter‘: 60000,
‘snapshot‘: 40000,
‘display‘: 10,
‘average_loss‘: 10,
‘type‘: "SGD",
‘solver_mode‘: solver_mode,
‘device_id‘: device_id,
‘debug_info‘: False,
‘snapshot_after_train‘: True,
# Test parameters
‘test_iter‘: [test_iter],
‘test_interval‘: 10000,
‘eval_type‘: "detection",
‘ap_version‘: "11point",
‘test_initialization‘: False,
}
# parameters for generating detection output.
det_out_param = {
‘num_classes‘: num_classes,
‘share_location‘: share_location,
‘background_label_id‘: background_label_id,
‘nms_param‘: {‘nms_threshold‘: 0.45, ‘top_k‘: 400},
‘save_output_param‘: {
‘output_directory‘: output_result_dir,
‘output_name_prefix‘: "comp4_det_test_",
‘output_format‘: "VOC",
‘label_map_file‘: label_map_file,
‘name_size_file‘: name_size_file,
‘num_test_image‘: num_test_image,
},
‘keep_top_k‘: 200,
‘confidence_threshold‘: 0.01,
‘code_type‘: code_type,
}
# parameters for evaluating detection results.
det_eval_param = {
‘num_classes‘: num_classes,
‘background_label_id‘: background_label_id,
‘overlap_threshold‘: 0.5,
‘evaluate_difficult_gt‘: False,
‘name_size_file‘: name_size_file,
}
### Hopefully you don‘t need to change the following ###
# Check file.
check_if_exist(train_data)
check_if_exist(test_data)
check_if_exist(label_map_file)
check_if_exist(pretrain_model)
make_if_not_exist(save_dir)
make_if_not_exist(job_dir)
make_if_not_exist(snapshot_dir)
# Create train net.
net = caffe.NetSpec()
net.data, net.label = CreateAnnotatedDataLayer(train_data, batch_size=batch_size_per_device,
train=True, output_label=True, label_map_file=label_map_file,
transform_param=train_transform_param, batch_sampler=batch_sampler)
VGGNetBody(net, from_layer=‘data‘, fully_conv=True, reduced=True, dilated=True,
dropout=False, freeze_layers=freeze_layers)
AddExtraLayers(net, use_batchnorm)
mbox_layers = CreateMultiBoxHead(net, data_layer=‘data‘, from_layers=mbox_source_layers,
use_batchnorm=use_batchnorm, min_sizes=min_sizes, max_sizes=max_sizes,
aspect_ratios=aspect_ratios, normalizations=normalizations,
num_classes=num_classes, share_location=share_location, flip=flip, clip=clip,
prior_variance=prior_variance, kernel_size=3, pad=1)
# Create the MultiBoxLossLayer.
name = "mbox_loss"
mbox_layers.append(net.label)
net[name] = L.MultiBoxLoss(*mbox_layers, multibox_loss_param=multibox_loss_param,
loss_param=loss_param, include=dict(phase=caffe_pb2.Phase.Value(‘TRAIN‘)),
propagate_down=[True, True, False, False])
with open(train_net_file, ‘w‘) as f:
print(‘name: "{}_train"‘.format(model_name), file=f)
print(net.to_proto(), file=f)
shutil.copy(train_net_file, job_dir)
# Create test net.
net = caffe.NetSpec()
net.data, net.label = CreateAnnotatedDataLayer(test_data, batch_size=test_batch_size,
train=False, output_label=True, label_map_file=label_map_file,
transform_param=test_transform_param)
VGGNetBody(net, from_layer=‘data‘, fully_conv=True, reduced=True, dilated=True,
dropout=False, freeze_layers=freeze_layers)
AddExtraLayers(net, use_batchnorm)
mbox_layers = CreateMultiBoxHead(net, data_layer=‘data‘, from_layers=mbox_source_layers,
use_batchnorm=use_batchnorm, min_sizes=min_sizes, max_sizes=max_sizes,
aspect_ratios=aspect_ratios, normalizations=normalizations,
num_classes=num_classes, share_location=share_location, flip=flip, clip=clip,
prior_variance=prior_variance, kernel_size=3, pad=1)
conf_name = "mbox_conf"
if multibox_loss_param["conf_loss_type"] == P.MultiBoxLoss.SOFTMAX:
reshape_name = "{}_reshape".format(conf_name)
net[reshape_name] = L.Reshape(net[conf_name], shape=dict(dim=[0, -1, num_classes]))
softmax_name = "{}_softmax".format(conf_name)
net[softmax_name] = L.Softmax(net[reshape_name], axis=2)
flatten_name = "{}_flatten".format(conf_name)
net[flatten_name] = L.Flatten(net[softmax_name], axis=1)
mbox_layers[1] = net[flatten_name]
elif multibox_loss_param["conf_loss_type"] == P.MultiBoxLoss.LOGISTIC:
sigmoid_name = "{}_sigmoid".format(conf_name)
net[sigmoid_name] = L.Sigmoid(net[conf_name])
mbox_layers[1] = net[sigmoid_name]
net.detection_out = L.DetectionOutput(*mbox_layers,
detection_output_param=det_out_param,
include=dict(phase=caffe_pb2.Phase.Value(‘TEST‘)))
net.detection_eval = L.DetectionEvaluate(net.detection_out, net.label,
detection_evaluate_param=det_eval_param,
include=dict(phase=caffe_pb2.Phase.Value(‘TEST‘)))
with open(test_net_file, ‘w‘) as f:
print(‘name: "{}_test"‘.format(model_name), file=f)
print(net.to_proto(), file=f)
shutil.copy(test_net_file, job_dir)
# Create deploy net.
# Remove the first and last layer from test net.
deploy_net = net
with open(deploy_net_file, ‘w‘) as f:
net_param = deploy_net.to_proto()
# Remove the first (AnnotatedData) and last (DetectionEvaluate) layer from test net.
del net_param.layer[0]
del net_param.layer[-1]
net_param.name = ‘{}_deploy‘.format(model_name)
net_param.input.extend([‘data‘])
net_param.input_shape.extend([
caffe_pb2.BlobShape(dim=[1, 3, resize_height, resize_width])])
print(net_param, file=f)
shutil.copy(deploy_net_file, job_dir)
# Create solver.
solver = caffe_pb2.SolverParameter(
train_net=train_net_file,
test_net=[test_net_file],
snapshot_prefix=snapshot_prefix,
**solver_param)
with open(solver_file, ‘w‘) as f:
print(solver, file=f)
shutil.copy(solver_file, job_dir)
max_iter = 0
# Find most recent snapshot.
for file in os.listdir(snapshot_dir):
if file.endswith(".solverstate"):
basename = os.path.splitext(file)[0]
iter = int(basename.split("{}_iter_".format(model_name))[1])
if iter > max_iter:
max_iter = iter
train_src_param = ‘--weights="{}" \\\n‘.format(pretrain_model)
if resume_training:
if max_iter > 0:
train_src_param = ‘--snapshot="{}_iter_{}.solverstate" \\\n‘.format(snapshot_prefix, max_iter)
if remove_old_models:
# Remove any snapshots smaller than max_iter.
for file in os.listdir(snapshot_dir):
if file.endswith(".solverstate"):
basename = os.path.splitext(file)[0]
iter = int(basename.split("{}_iter_".format(model_name))[1])
if max_iter > iter:
os.remove("{}/{}".format(snapshot_dir, file))
if file.endswith(".caffemodel"):
basename = os.path.splitext(file)[0]
iter = int(basename.split("{}_iter_".format(model_name))[1])
if max_iter > iter:
os.remove("{}/{}".format(snapshot_dir, file))
# Create job file.
with open(job_file, ‘w‘) as f:
f.write(‘cd {}\n‘.format(caffe_root))
f.write(‘./build/tools/caffe train \\\n‘)
f.write(‘--solver="{}" \\\n‘.format(solver_file))
f.write(train_src_param)
if solver_param[‘solver_mode‘] == P.Solver.GPU:
f.write(‘--gpu {} 2>&1 | tee {}/{}.log\n‘.format(gpus, job_dir, model_name))
else:
f.write(‘2>&1 | tee {}/{}.log\n‘.format(job_dir, model_name))
# Copy the python script to job_dir.
py_file = os.path.abspath(__file__)
shutil.copy(py_file, job_dir)
# Run the job.
os.chmod(job_file, stat.S_IRWXU)
if run_soon:
subprocess.call(job_file, shell=True)
ssd_pascal_indoor.py
训练命令：python examples/ssd/ssd_pascal_indoor.py
未完待续......
时间： 2024-11-08 22:47:50