Wei Liu大神SSD工程：https://github.com/weiliu89/caffe/tree/ssd
1.编译caffe，下载预训练模型存放于./caffe/models/VGGNet
https://pan.baidu.com/s/1oYquStMO-c70hLtwEXzD8Q
2.在./caffe/data下新建VOCdevkit文件夹，VOCdevkit下再建一个自己的数据集文件夹，文件夹下为Annotations、ImageSets、JPEGImages三个训练所需数据文件夹。
3.在./caffe/data下再建一个自己的数据集文件夹，从./data/VOC0712中复制create_data.sh、create_lish.sh、labelmap_voc.prototxt三个文件过来。
修改create_lish.sh
#!/bin/bash
root_dir=/xxx/xxx/SSD/caffe/data/VOCdevkit ##########修改为VOCdevkit 所在的绝对路径
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 xxx ####################修改为自己建立的数据文件夹名，不含中文路径
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 $dst_file
fi
done
执行完之后得到test.txt、test_name_size.txt、trainval.txt三个txt文件。
修改create_data.sh
cur_dir=$(cd $( dirname ${BASH_SOURCE[0]} ) && pwd )
root_dir=$cur_dir/../..
cd $root_dir
redo=1
data_root_dir="/xxx/xxx/SSD/caffe/data/VOCdevkit" ##########修改为VOCdevkit 所在的绝对路径
dataset_name="xxx" ####################修改为自己建立的数据文件夹名，不含中文路径
mapfile="$root_dir/data/$dataset_name/labelmap_voc.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
根目录执行后在caffe/data/VOCdevkit/自己建立的数据文件夹名/ 下得到二进制lmdb文件夹
修改labelmap_voc.prototxt
item {
name: "none_of_the_above"
label: 0
display_name: "background"
}
item {
name: "person"
label: 1
display_name: "person" ###########我只有一类person
}
4.修改./caffe/examples/ssd/ssd_pascal.py
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, lr_mult=1):
use_relu = True
# Add additional convolutional layers.
# 19 x 19
from_layer = net.keys()[-1]
# TODO(weiliu89): Construct the name using the last layer to avoid duplication.
# 10 x 10
out_layer = "conv6_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv6_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 512, 3, 1, 2,
lr_mult=lr_mult)
# 5 x 5
from_layer = out_layer
out_layer = "conv7_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv7_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 1, 2,
lr_mult=lr_mult)
# 3 x 3
from_layer = out_layer
out_layer = "conv8_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv8_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 0, 1,
lr_mult=lr_mult)
# 1 x 1
from_layer = out_layer
out_layer = "conv9_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv9_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 0, 1,
lr_mult=lr_mult)
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/xxx/xxx_trainval_lmdb" ################修改二进制数据路径
# The database file for testing data. Created by data/VOC0712/create_data.sh
test_data = "examples/xxx/xxxx_test_lmdb" ###############修改二进制数据路径
# Specify the batch sampler.
resize_width = 300#300
resize_height = 300#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': [88, 86, 88],#104,117,123 ##########可以用build下的compute_image_mean计算一下自己这批数据的均值
'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,
],
},
'distort_param': {
'brightness_prob': 0.5,
'brightness_delta': 32,
'contrast_prob': 0.5,
'contrast_lower': 0.5,
'contrast_upper': 1.5,
'hue_prob': 0.5,
'hue_delta': 18,
'saturation_prob': 0.5,
'saturation_lower': 0.5,
'saturation_upper': 1.5,
'random_order_prob': 0.0,
},
'expand_param': {
'prob': 0.5,
'max_expand_ratio': 4.0,
},
'emit_constraint': {
'emit_type': caffe_pb2.EmitConstraint.CENTER,
}
}
test_transform_param = {
'mean_value': [88, 86, 88],#104,117,123 ############修改均值
'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
lr_mult = 1
# 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_xxx_{}".format(job_name) #############xxx为自己建立的数据文件夹名
# Directory which stores the model .prototxt file.
save_dir = "models/VGGNet/xxx/{}".format(job_name) ############xxx为自己建立的数据文件夹名
# Directory which stores the snapshot of models.
snapshot_dir = "models/VGGNet/xxx/{}".format(job_name) ###########xxx为自己建立的数据文件夹名
# Directory which stores the job script and log file.
job_dir = "jobs/VGGNet/xxx/{}".format(job_name) ############xxx为自己建立的数据文件夹名
# Directory which stores the detection results.
output_result_dir = "{}/SSD/caffe/data/VOCdevkit/results/xxx/{}/Main".format(os.environ['HOME'], job_name) ########xxx为自己建立的数据文件夹名
# 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/xxx/test_name_size.txt" ############xxx为自己建立的数据文件夹名
# The pretrained model. We use the Fully convolutional reduced (atrous) VGGNet.
pretrain_model = "models/VGGNet/VGG_ILSVRC_16_layers_fc_reduced.caffemodel"##########预训练模型
# Stores LabelMapItem.
label_map_file = "data/xxx/labelmap_voc.prototxt" ##########xxx为自己建立的数据文件夹名
# MultiBoxLoss parameters.
num_classes = 2 ####################类别数为自己的类别数+1
share_location = True
background_label_id=0
train_on_diff_gt = True
normalization_mode = P.Loss.VALID
code_type = P.PriorBox.CENTER_SIZE
ignore_cross_boundary_bbox = False
mining_type = P.MultiBoxLoss.MAX_NEGATIVE
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,
'mining_type': mining_type,
'neg_pos_ratio': neg_pos_ratio,
'neg_overlap': 0.5,
'code_type': code_type,
'ignore_cross_boundary_bbox': ignore_cross_boundary_bbox,
}
loss_param = {
'normalization': normalization_mode,
}
# parameters for generating priors.
# minimum dimension of input image
min_dim = 300#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
# conv9_2 ==> 1 x 1
mbox_source_layers = ['conv4_3', 'fc7', 'conv6_2', 'conv7_2', 'conv8_2', 'conv9_2']
# in percent %
min_ratio = 20
max_ratio = 90
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 = [min_dim * 20 / 100.] + max_sizes
steps = [8, 16, 32, 64, 100, 300]
aspect_ratios = [[2], [2, 3], [2, 3], [2, 3], [2], [2]]
# 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 = False
# Solver parameters.
# Defining which GPUs to use.
gpus = "1" ###########根据自己的显卡索引号修改
gpulist = gpus.split(",")
num_gpus = len(gpulist)
# Divide the mini-batch to different GPUs.
batch_size = 32
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.
# Evaluate on whole test set.
num_test_image = 7716 ################test.txt的行数
test_batch_size = 8
# Ideally test_batch_size should be divisible by num_test_image,
# otherwise mAP will be slightly off the true value.
test_iter = int(math.ceil(float(num_test_image) / test_batch_size))
solver_param = {
# Train parameters
'base_lr': base_lr,
'weight_decay': 0.0005,
'lr_policy': "multistep",
'stepvalue': [40000, 60000, 80000],#80000,100000,120000
'gamma': 0.1,
'momentum': 0.9,
'iter_size': iter_size,
'max_iter': 80000,#120000
'snapshot': 10000,#80000
'display': 500,
'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': 500,#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)
AddExtraLayers(net, use_batchnorm, lr_mult=lr_mult)
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, steps=steps, normalizations=normalizations,
num_classes=num_classes, share_location=share_location, flip=flip, clip=clip,
prior_variance=prior_variance, kernel_size=3, pad=1, lr_mult=lr_mult)
# 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)
AddExtraLayers(net, use_batchnorm, lr_mult=lr_mult)
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, steps=steps, normalizations=normalizations,
num_classes=num_classes, share_location=share_location, flip=flip, clip=clip,
prior_variance=prior_variance, kernel_size=3, pad=1, lr_mult=lr_mult)
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)
根目录执行，开始训练
扫描二维码关注公众号，回复：
148481 查看本文章
python ./examples/ssd/ssd_pascal.py
5.测试图片
修改./caffe/examples/ssd/ssd_detect.py
#encoding=utf8
'''
Detection with SSD
In this example, we will load a SSD model and use it to detect objects.
'''
import os
import sys
import argparse
import numpy as np
from PIL import Image, ImageDraw
# Make sure that caffe is on the python path:
caffe_root = './'
os.chdir(caffe_root)
sys.path.insert(0, os.path.join(caffe_root, 'python'))
import caffe
from google.protobuf import text_format
from caffe.proto import caffe_pb2
def get_labelname(labelmap, labels):
num_labels = len(labelmap.item)
labelnames = []
if type(labels) is not list:
labels = [labels]
for label in labels:
found = False
for i in xrange(0, num_labels):
if label == labelmap.item[i].label:
found = True
labelnames.append(labelmap.item[i].display_name)
break
assert found == True
return labelnames
class CaffeDetection:
def __init__(self, gpu_id, model_def, model_weights, image_resize, labelmap_file):
caffe.set_device(gpu_id)
caffe.set_mode_gpu()
self.image_resize = image_resize
# Load the net in the test phase for inference, and configure input preprocessing.
self.net = caffe.Net(model_def, # defines the structure of the model
model_weights, # contains the trained weights
caffe.TEST) # use test mode (e.g., don't perform dropout)
# input preprocessing: 'data' is the name of the input blob == net.inputs[0]
self.transformer = caffe.io.Transformer({'data': self.net.blobs['data'].data.shape})
self.transformer.set_transpose('data', (2, 0, 1))
self.transformer.set_mean('data', np.array([88, 86, 88])) # mean pixel ##################修改均值
# the reference model operates on images in [0,255] range instead of [0,1]
self.transformer.set_raw_scale('data', 255)
# the reference model has channels in BGR order instead of RGB
self.transformer.set_channel_swap('data', (2, 1, 0))
# load PASCAL VOC labels
file = open(labelmap_file, 'r')
self.labelmap = caffe_pb2.LabelMap()
text_format.Merge(str(file.read()), self.labelmap)
def detect(self, image_file, conf_thresh=0.5, topn=5):
'''
SSD detection
'''
# set net to batch size of 1
# image_resize = 300
self.net.blobs['data'].reshape(1, 3, self.image_resize, self.image_resize)
image = caffe.io.load_image(image_file)
#Run the net and examine the top_k results
transformed_image = self.transformer.preprocess('data', image)
self.net.blobs['data'].data[...] = transformed_image
# Forward pass.
detections = self.net.forward()['detection_out']
# Parse the outputs.
det_label = detections[0,0,:,1]
det_conf = detections[0,0,:,2]
det_xmin = detections[0,0,:,3]
det_ymin = detections[0,0,:,4]
det_xmax = detections[0,0,:,5]
det_ymax = detections[0,0,:,6]
# Get detections with confidence higher than 0.6.
top_indices = [i for i, conf in enumerate(det_conf) if conf >= conf_thresh]
top_conf = det_conf[top_indices]
top_label_indices = det_label[top_indices].tolist()
top_labels = get_labelname(self.labelmap, top_label_indices)
top_xmin = det_xmin[top_indices]
top_ymin = det_ymin[top_indices]
top_xmax = det_xmax[top_indices]
top_ymax = det_ymax[top_indices]
result = []
for i in xrange(min(topn, top_conf.shape[0])):
xmin = top_xmin[i] # xmin = int(round(top_xmin[i] * image.shape[1]))
ymin = top_ymin[i] # ymin = int(round(top_ymin[i] * image.shape[0]))
xmax = top_xmax[i] # xmax = int(round(top_xmax[i] * image.shape[1]))
ymax = top_ymax[i] # ymax = int(round(top_ymax[i] * image.shape[0]))
score = top_conf[i]
label = int(top_label_indices[i])
label_name = top_labels[i]
result.append([xmin, ymin, xmax, ymax, label, score, label_name])
return result
def main(args):
'''main '''
detection = CaffeDetection(args.gpu_id,
args.model_def, args.model_weights,
args.image_resize, args.labelmap_file)
result = detection.detect(args.image_file)
print result
img = Image.open(args.image_file)
draw = ImageDraw.Draw(img)
width, height = img.size
print width, height
for item in result:
xmin = int(round(item[0] * width))
ymin = int(round(item[1] * height))
xmax = int(round(item[2] * width))
ymax = int(round(item[3] * height))
draw.rectangle([xmin, ymin, xmax, ymax], outline=(255, 0, 0))
draw.text([xmin, ymin], item[-1] + str(item[-2]), (0, 0, 255))
print item
print [xmin, ymin, xmax, ymax]
print [xmin, ymin], item[-1]
img.save('detect_result.jpg')
def parse_args():
'''parse args'''
parser = argparse.ArgumentParser()
parser.add_argument('--gpu_id', type=int, default=0, help='gpu id')
parser.add_argument('--labelmap_file',
default='data/xxx/labelmap_voc.prototxt') ############自己的标签文件
parser.add_argument('--model_def',
default='models/VGGNet/xxx/SSD_300x300/deploy.prototxt') ##########训练后在models/VGGNet下会生成测试网络
parser.add_argument('--image_resize', default=300, type=int)
parser.add_argument('--model_weights',
default='models/VGGNet/xxx/SSD_300x300/'
'VGG_elevator_SSD_300x300_iter_80000.caffemodel') #############训练得到的模型
parser.add_argument('--image_file', default='examples/images/8.jpg') #########测试的图片名
return parser.parse_args()
if __name__ == '__main__':
main(parse_args())
根目录执行
python ./examples/ssd/ssd_detect.py
6.测试视频
修改./caffe/examples/ssd/ssd_pascal_video.py
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, lr_mult=1):
use_relu = True
# Add additional convolutional layers.
# 19 x 19
from_layer = net.keys()[-1]
# TODO(weiliu89): Construct the name using the last layer to avoid duplication.
# 10 x 10
out_layer = "conv6_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv6_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 512, 3, 1, 2,
lr_mult=lr_mult)
# 5 x 5
from_layer = out_layer
out_layer = "conv7_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv7_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 1, 2,
lr_mult=lr_mult)
# 3 x 3
from_layer = out_layer
out_layer = "conv8_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv8_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 0, 1,
lr_mult=lr_mult)
# 1 x 1
from_layer = out_layer
out_layer = "conv9_1"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 128, 1, 0, 1,
lr_mult=lr_mult)
from_layer = out_layer
out_layer = "conv9_2"
ConvBNLayer(net, from_layer, out_layer, use_batchnorm, use_relu, 256, 3, 0, 1,
lr_mult=lr_mult)
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
# The video file path
video_file = "examples/videos/1.mp4" ############测试的视频文件
# The parameters for the video demo
# Key parameters used in training
# If true, use batch norm for all newly added layers.
# Currently only the non batch norm version has been tested.
use_batchnorm = False
num_classes = 2 ###########自己的类别数+1
share_location = True
background_label_id=0
conf_loss_type = P.MultiBoxLoss.SOFTMAX
code_type = P.PriorBox.CENTER_SIZE
lr_mult = 1.
# Stores LabelMapItem.
label_map_file = "data/xxx/labelmap_voc.prototxt" ###########自己新建的数据集文件夹名
# The resized image size
resize_width = 300
resize_height = 300
# Parameters needed for test.
# Set the number of test iterations to the maximum integer number.
test_iter = int(math.pow(2, 29) - 1)
# Use GPU or CPU
solver_mode = P.Solver.GPU
# Defining which GPUs to use.
gpus = "0"
# Number of frames to be processed per batch.
test_batch_size = 1
# Only display high quality detections whose scores are higher than a threshold.
visualize_threshold = 0.3
# Size of video image.
video_width = 576 ##############视频文件的宽高
video_height = 704 ##############视频文件的宽高
# Scale the image size for display.
scale = 0.8
### Hopefully you don't need to change the following ###
resize = "{}x{}".format(resize_width, resize_height)
video_data_param = {
'video_type': P.VideoData.VIDEO,
'video_file': video_file,
}
test_transform_param = {
'mean_value': [88, 86, 88], ##########修改均值
'resize_param': {
'prob': 1,
'resize_mode': P.Resize.WARP,
'height': resize_height,
'width': resize_width,
'interp_mode': [P.Resize.LINEAR],
},
}
output_transform_param = {
'mean_value': [88, 86, 88], ##########修改均值
'resize_param': {
'prob': 1,
'resize_mode': P.Resize.WARP,
'height': int(video_height * scale),
'width': int(video_width * scale),
'interp_mode': [P.Resize.LINEAR],
},
}
# 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': {
'label_map_file': label_map_file,
},
'keep_top_k': 200,
'confidence_threshold': 0.01,
'code_type': code_type,
'visualize': True,
'visualize_threshold': visualize_threshold,
}
# The job name should be same as the name used in examples/ssd/ssd_pascal.py.
job_name = "SSD_{}".format(resize)
# The name of the model. Modify it if you want.
model_name = "VGG_xxx_{}".format(job_name) ############xxx自己新建的数据集文件夹名
# Directory which stores the model .prototxt file.
save_dir = "models/xxx/elevator/{}_video".format(job_name) ###########xxx自己新建的数据集文件夹名
# Directory which stores the snapshot of trained models.
snapshot_dir = "models/xxx/elevator/{}".format(job_name) #############xxx自己新建的数据集文件夹名
# Directory which stores the job script and log file.
job_dir = "jobs/VGGNet/xxx/{}_video".format(job_name) ############xxx自己新建的数据集文件夹名
# model definition files.
test_net_file = "{}/test.prototxt".format(save_dir)
# snapshot prefix.
snapshot_prefix = "{}/{}".format(snapshot_dir, model_name)
# job script path.
job_file = "{}/{}.sh".format(job_dir, model_name)
# Find most recent snapshot.
max_iter = 0
for file in os.listdir(snapshot_dir):
if file.endswith(".caffemodel"):
basename = os.path.splitext(file)[0]
iter = int(basename.split("{}_iter_".format(model_name))[1])
if iter > max_iter:
max_iter = iter
if max_iter == 0:
print("Cannot find snapshot in {}".format(snapshot_dir))
sys.exit()
# The resume model.
pretrain_model = "{}_iter_{}.caffemodel".format(snapshot_prefix, max_iter)
# 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
# conv9_2 ==> 1 x 1
mbox_source_layers = ['conv4_3', 'fc7', 'conv6_2', 'conv7_2', 'conv8_2', 'conv9_2']
# in percent %
min_ratio = 20
max_ratio = 90
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 = [min_dim * 20 / 100.] + max_sizes
steps = [8, 16, 32, 64, 100, 300]
aspect_ratios = [[2], [2, 3], [2, 3], [2, 3], [2], [2]]
# 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 = False
# Check file.
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 test net.
net = caffe.NetSpec()
net.data = L.VideoData(video_data_param=video_data_param,
data_param=dict(batch_size=test_batch_size),
transform_param=test_transform_param)
VGGNetBody(net, from_layer='data', fully_conv=True, reduced=True, dilated=True,
dropout=False)
AddExtraLayers(net, use_batchnorm, lr_mult=lr_mult)
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, steps=steps, normalizations=normalizations,
num_classes=num_classes, share_location=share_location, flip=flip, clip=clip,
prior_variance=prior_variance, kernel_size=3, pad=1, lr_mult=lr_mult)
conf_name = "mbox_conf"
if 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 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]
mbox_layers.append(net.data)
net.detection_out = L.DetectionOutput(*mbox_layers,
detection_output_param=det_out_param,
transform_param=output_transform_param,
include=dict(phase=caffe_pb2.Phase.Value('TEST')))
net.slience = L.Silence(net.detection_out, ntop=0,
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 job file.
with open(job_file, 'w') as f:
f.write('cd {}\n'.format(caffe_root))
f.write('./build/tools/caffe test \\\n')
f.write('--model="{}" \\\n'.format(test_net_file))
f.write('--weights="{}" \\\n'.format(pretrain_model))
f.write('--iterations="{}" \\\n'.format(test_iter))
if solver_mode == P.Solver.GPU:
f.write('--gpu {}\n'.format(gpus))
# 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)
根目录下执行
python ./examples/ssd/ssd_pascal_video.py
7.SSD的mAP计算
训练日志文件保存在./caffe/jobs/VGGNet/xxx/SSD_300x300文件夹下，ssd_pascal.py中solver_param下有一个 'eval_type': "detection",因此日志文件中有Test net output #0: detection_eval即mAP值，测试一批数据的mAP需将test.txt换成这批数据做训练即可。