Deeplab v2 调试全过程（Ubuntu 16.04+cuda8.0）

Deeplab v2 调试全过程（Ubuntu 16.04+cuda8.0）

本人刚接触深度学习与caffe，经过几天的填坑，终于把Deeplabv2的 run_pascal.sh与run_densecrf.sh成功运行，现将调试过程整理如下：

• 一、安装必要的依赖库
  安装 matio:
  安装方法1：
  sudo apt-get install libmatio-dev
  安装方法2：
  下载matio(https://sourceforge.net/projects/matio/files/matio/1.5.2/)
  tar zxf matio-1.5.2.tar.gz
  cd matio-1.5.2
  ./configure
  make
  make check
  make install
  sudo ldconfig
  安装 wget
  sudo pip install wget 出错
  按照下面的命令成功：
  pip install –upgrade pip –user
  pip install –upgrade setuptools –user
  sudo pip install wget

• 二、下载Deeplabv2并编译
  1、下载代码：
  git clone https://github.com/xmojiao/deeplab_v2.git
  （试过许多Deeplab代码，这个最容易编译成功，所以我用的是这个代码编译的）。
  2、对 caffe 进行编译：
  修改deeplab_v2/deeplab-public-ver2/路径下的 Makefile.config.example文件，重命名为Makefile.config，接着修改这个文件中的内容，将第四行的 “# USE_CUDNN := 1”的 # 去掉。如果需要，因为我用的pycaffe编译，所以不需要修改python的路径，保存退出。
  下面为编译 caffe的命令：
  cd ~/Desktop/deeplab_v2/deeplab-public-ver2
  make all -j16
  这时会出现下面的错误1：
  src/caffe/net.cpp:8:18: fatal error: hdf5.h: No such file or directory
  compilation terminated.
  解决办法: 修改两个make文件（Makefile.config，Makefile）
  Makefile.config：
  将
  INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib
  修改为：
  INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnumake
  Makefile：
  将
  LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_hl hdf5
  修改为：
  LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial matio
  重新编译：
  make all -j16
  这时会出现下面的错误2：
  ./include/caffe/common.cuh(9): error: function “atomicAdd(double *, double)” has already been defined
  **解决方法：打开./include/caffe/common.cuh文件，在atomicAdd前添加宏判断即可。
  下面为修改后文件：**

   // Copyright 2014 George Papandreou

    #ifndef CAFFE_COMMON_CUH_
    #define CAFFE_COMMON_CUH_

    #include 


    // CUDA: atomicAdd is not defined for doubles

    #if !defined(__CUDA_ARCH__) || __CUDA_ARCH__ >= 600 
    #else
    static __inline__ __device__ double atomicAdd(double *address, double val) {
      unsigned long long int* address_as_ull = (unsigned long long int*)address;
      unsigned long long int old = *address_as_ull, assumed;
      if (val==0.0)
        return __longlong_as_double(old);
      do {
        assumed = old;
        old = atomicCAS(address_as_ull, assumed, __double_as_longlong(val +__longlong_as_double(assumed)));
      } while (assumed != old);
      return __longlong_as_double(old);
    }

    #endif
    #endif

继续编译： make all -j16

这时会出现下面的错误3：:.build_release/lib/libcaffe.so:undefined
reference to `cudnnConvolutionBackwardFilter_v3’
解决方法：
将BVLC(https://github.com/BVLC/caffe)中的下列文件copy 到相应的文件夹：
./include/caffe/util/cudnn.hpp
./include/caffe/layers/cudnn_conv_layer.hpp
./include/caffe/layers/cudnn_relu_layer.hpp
./include/caffe/layers/cudnn_sigmoid_layer.hpp
./include/caffe/layers/cudnn_tanh_layer.hpp
./src/caffe/layers/cudnn_conv_layer.cpp
./src/caffe/layers/cudnn_conv_layer.cu
./src/caffe/layers/cudnn_relu_layer.cpp
./src/caffe/layers/cudnn_relu_layer.cu
./src/caffe/layers/cudnn_sigmoid_layer.cpp
./src/caffe/layers/cudnn_sigmoid_layer.cu
./src/caffe/layers/cudnn_tanh_layer.cpp
./src/caffe/layers/cudnn_tanh_layer.cu

make clean
make all -j16
make pycaffe -j16
编译成功。

2、对 run_pascal.sh 进行调试：
（1）首先准备好数据
（我是按照这篇博客准备的数据： http://blog.csdn.net/Xmo_jiao/article/details/77897109）
cd ~/Desktop
mkdir -p my_dataset
# augmented PASCAL VOC
cd my_dataset/
wget http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/semantic_contours/benchmark.tgz # 1.3 GB
tar -zxvf benchmark.tgz
mv benchmark_RELEASE VOC_aug

# original PASCAL VOC 2012
wget http://host.robots.ox.ac.uk/pascal/VOC/voc2012/VOCtrainval_11-May-2012.tar # 2 GB
tar -xvf VOCtrainval_11-May-2012.tar
mv VOCdevkit/VOC2012 VOC2012_orig && rm -r VOCdevkit

(2)数据转换

因为pascal voc2012增强数据集的label是mat格式的文件，要把mat格式的label转为png格式的图片.
~/Desktop/my_dataset/VOC_aug/dataset
mkdir cls_png
cd ~/Desktop/deeplab_v2/voc2012/
./mat2png.py ~/Desktop/my_dataset/VOC_aug/dataset/cls /Desktop/my_dataset/VOC_aug/dataset/cls_png

因为pascal voc2012原始数据集的label为三通道RGB图像，但是caffe最后一层softmax loss
层只能识别一通道的label,所以此处我们需要对原始数据集的label进行降维
cd ~/Desktop/my_dataset/VOC2012_orig
mkdir SegmentationClass_1D
cd ~/Desktop/deeplab_v2/voc2012/
./convert_labels.py ~/Desktop/my_dataset/VOC2012_orig/SegmentationClass/ ~/Desktop/my_dataset
/VOC2012_orig/ImageSets/Segmentation/trainval.txt ~/Desktop/my_dataset/VOC2012_orig/Segmentat
ionClass_1D/

（3）数据融合

此时已经处理好好pascal voc2012 增强数据集和pascal voc2012的原始数据集，为了便于train.txt等文件的调用，将两个文件夹数据合并到同一个文件中.现有文件目录如下：

现分别pascal voc2012增强数据集里的images和labels复制到增强数据集中，若重复则覆盖，合将并数据集的操作如下：
cp ~/Desktop/my_dataset/VOC2012_orig/SegmentationClass_1D/* ~/Desktop/my_dataset/VOC_aug/dataset/cls_png
cp ~/Desktop/my_dataset/VOC2012_orig/JPEGImages/* ~/Desktop/my_dataset/VOC_aug/dataset/img/

(4)文件名修改

对应train.txt文件的数据集文件名，修改文件名。
cd ~/Desktop/my_dataset/VOC_aug/dataset
mv ./img ./JPEGImages
那么我们这个阶段使用的数据已经整理完成

（5）修改并运行 run_pascal.sh
1）准备必要的文件
需要的文件从这里下载 deeplabv2 有两种模型（vgg，Res102），我用的vgg ，http://liangchiehchen.com/projects/DeepLab_Models.html
总共需要的文件如图所示：

下载的代码中 Desktop/deeplab_v2/voc2012/list 已经有了list文件，所以不用重新下载。

/Desktop/deeplab_v2/voc2012/config/deeplab_largeFOV中也有了相应的文件，所以也无需下载。

Desktop/deeplab_v2/voc2012/model/deeplab_largeFOV 里没有model，需要把下载好的model放入文件，如图所示：

至此，所有需要的文件全部完毕。

2）运行 train 和 test
进入/Desktop/deeplab_v2/voc2012，修改 run_pascal.sh 文件，主要是修改路径，我的修改后的文件如下：

  #!/bin/sh

## MODIFY PATH for YOUR SETTING
ROOT_DIR=/home/mmt/Desktop/my_dataset

CAFFE_DIR=/home/mmt/Desktop/deeplab_v2/deeplab-public-ver2
CAFFE_BIN=${CAFFE_DIR}/build/tools/caffe.bin

EXP=.

if [ "${EXP}" = "." ]; then
    NUM_LABELS=21
    DATA_ROOT=${ROOT_DIR}/VOC_aug/dataset/
else
    NUM_LABELS=0
    echo "Wrong exp name"
fi


## Specify which model to train
########### voc12 ################
NET_ID=deeplab_largeFOV


## Variables used for weakly or semi-supervisedly training
#TRAIN_SET_SUFFIX=
TRAIN_SET_SUFFIX=_aug

#TRAIN_SET_STRONG=train
#TRAIN_SET_STRONG=train200
#TRAIN_SET_STRONG=train500
#TRAIN_SET_STRONG=train1000
#TRAIN_SET_STRONG=train750

#TRAIN_SET_WEAK_LEN=5000

DEV_ID=0

#####

## Create dirs

CONFIG_DIR=${EXP}/config/${NET_ID}
MODEL_DIR=${EXP}/model/${NET_ID}
mkdir -p ${MODEL_DIR}
LOG_DIR=${EXP}/log/${NET_ID}
mkdir -p ${LOG_DIR}
export GLOG_log_dir=${LOG_DIR}

## Run

RUN_TRAIN=1   #1时train
RUN_TEST=0    #1时test
RUN_TRAIN2=0
RUN_TEST2=0

## Training #1 (on train_aug)

if [ ${RUN_TRAIN} -eq 1 ]; then
    #
    LIST_DIR=${EXP}/list
    TRAIN_SET=train${TRAIN_SET_SUFFIX}
    if [ -z ${TRAIN_SET_WEAK_LEN} ]; then
                TRAIN_SET_WEAK=${TRAIN_SET}_diff_${TRAIN_SET_STRONG}
                comm -3 ${LIST_DIR}/${TRAIN_SET}.txt ${LIST_DIR}/${TRAIN_SET_STRONG}.txt > ${LIST_DIR}/${TRAIN_SET_WEAK}.txt
    else
                TRAIN_SET_WEAK=${TRAIN_SET}_diff_${TRAIN_SET_STRONG}_head${TRAIN_SET_WEAK_LEN}
                comm -3 ${LIST_DIR}/${TRAIN_SET}.txt ${LIST_DIR}/${TRAIN_SET_STRONG}.txt | head -n ${TRAIN_SET_WEAK_LEN} > ${LIST_DIR}/${TRAIN_SET_WEAK}.txt
    fi
    #
    MODEL=${EXP}/model/${NET_ID}/init.caffemodel
    #
    echo Training net ${EXP}/${NET_ID}
    for pname in train solver; do
                sed "$(eval echo $(cat sub.sed))" \
                        ${CONFIG_DIR}/${pname}.prototxt > ${CONFIG_DIR}/${pname}_${TRAIN_SET}.prototxt
    done
        CMD="${CAFFE_BIN} train \
         --solver=${CONFIG_DIR}/solver_${TRAIN_SET}.prototxt \
         --gpu=${DEV_ID}"
        if [ -f ${MODEL} ]; then
                CMD="${CMD} --weights=${MODEL}"
        fi
        echo Running ${CMD} && ${CMD}
fi

## Test #1 specification (on val or test)

if [ ${RUN_TEST} -eq 1 ]; then
    #
    for TEST_SET in val; do
                TEST_ITER=`cat ${EXP}/list/${TEST_SET}.txt | wc -l`
                MODEL=${EXP}/model/${NET_ID}/test.caffemodel
                if [ ! -f ${MODEL} ]; then
                        MODEL=`ls -t ${EXP}/model/${NET_ID}/train_iter_*.caffemodel | head -n 1`
                fi
                #
                echo Testing net ${EXP}/${NET_ID}
                FEATURE_DIR=${EXP}/features/${NET_ID}
                mkdir -p ${FEATURE_DIR}/${TEST_SET}/fc8
        mkdir -p ${FEATURE_DIR}/${TEST_SET}/fc9
                mkdir -p ${FEATURE_DIR}/${TEST_SET}/seg_score
                sed "$(eval echo $(cat sub.sed))" \
                        ${CONFIG_DIR}/test.prototxt > ${CONFIG_DIR}/test_${TEST_SET}.prototxt
                CMD="${CAFFE_BIN} test \
             --model=${CONFIG_DIR}/test_${TEST_SET}.prototxt \
             --weights=${MODEL} \
             --gpu=${DEV_ID} \
             --iterations=${TEST_ITER}"
                echo Running ${CMD} && ${CMD}
    done
fi

## Training #2 (finetune on trainval_aug)

if [ ${RUN_TRAIN2} -eq 1 ]; then
    #
    LIST_DIR=${EXP}/list
    TRAIN_SET=trainval${TRAIN_SET_SUFFIX}
    if [ -z ${TRAIN_SET_WEAK_LEN} ]; then
                TRAIN_SET_WEAK=${TRAIN_SET}_diff_${TRAIN_SET_STRONG}
                comm -3 ${LIST_DIR}/${TRAIN_SET}.txt ${LIST_DIR}/${TRAIN_SET_STRONG}.txt > ${LIST_DIR}/${TRAIN_SET_WEAK}.txt
    else
                TRAIN_SET_WEAK=${TRAIN_SET}_diff_${TRAIN_SET_STRONG}_head${TRAIN_SET_WEAK_LEN}
                comm -3 ${LIST_DIR}/${TRAIN_SET}.txt ${LIST_DIR}/${TRAIN_SET_STRONG}.txt | head -n ${TRAIN_SET_WEAK_LEN} > ${LIST_DIR}/${TRAIN_SET_WEAK}.txt
    fi
    #
    MODEL=${EXP}/model/${NET_ID}/init2.caffemodel
    if [ ! -f ${MODEL} ]; then
                MODEL=`ls -t ${EXP}/model/${NET_ID}/train_iter_*.caffemodel | head -n 1`
    fi
    #
    echo Training2 net ${EXP}/${NET_ID}
    for pname in train solver2; do
                sed "$(eval echo $(cat sub.sed))" \
                        ${CONFIG_DIR}/${pname}.prototxt > ${CONFIG_DIR}/${pname}_${TRAIN_SET}.prototxt
    done
    CMD="${CAFFE_BIN} train \
         --solver=${CONFIG_DIR}/solver2_${TRAIN_SET}.prototxt \
         --weights=${MODEL} \
         --gpu=${DEV_ID}"
        echo Running ${CMD} && ${CMD}
fi

## Test #2 on official test set

if [ ${RUN_TEST2} -eq 1 ]; then
    #
    for TEST_SET in val test; do
                TEST_ITER=`cat ${EXP}/list/${TEST_SET}.txt | wc -l`
                MODEL=${EXP}/model/${NET_ID}/test2.caffemodel
                if [ ! -f ${MODEL} ]; then
                        MODEL=`ls -t ${EXP}/model/${NET_ID}/train2_iter_*.caffemodel | head -n 1`
                fi
                #
                echo Testing2 net ${EXP}/${NET_ID}
                FEATURE_DIR=${EXP}/features2/${NET_ID}
                mkdir -p ${FEATURE_DIR}/${TEST_SET}/fc8
                mkdir -p ${FEATURE_DIR}/${TEST_SET}/crf
                sed "$(eval echo $(cat sub.sed))" \
                        ${CONFIG_DIR}/test.prototxt > ${CONFIG_DIR}/test_${TEST_SET}.prototxt
                CMD="${CAFFE_BIN} test \
             --model=${CONFIG_DIR}/test_${TEST_SET}.prototxt \
             --weights=${MODEL} \
             --gpu=${DEV_ID} \
             --iterations=${TEST_ITER}"
                echo Running ${CMD} && ${CMD}
    done
fi

接下来运行代码：
Train：
~/Desktop/deeplab_v2/voc2012
sh ./run_pascal.sh
运行结果如下：

Test：
将相应变量改为1：
~/Desktop/deeplab_v2/voc2012
sh ./run_pascal.sh
运行结果如下：

因为结果保存的是mat文件，如果想转换成png的形式，运行：
cd ~/Desktop/deeplab_v2/voc2012
修改create_labels_21.py的路径，然后此目录运行：
python create_labels_21.py
因为训练一会，我就暂停了，所以test的结果不好，而且图像的分割后的尺寸变了，不知道什么原因，不过经过densecrf后会变回来。

（6）修改并运行 run_densecrf.sh
1）首先对densecrf进行编译。
cd ~/Desktop/deeplab_v2/deeplab-public-ver2/densecrf/
make
有许多warning，但是没出错，不用管。
2）数据整理
因为densecrf只识别ppm格式的图像，所以要转换格式。进入/Desktop/deeplab_v2/deeplab-public-ver2/densecrf/my_script，里面有自带的修改ppm 的MATLAB程序，修改路径，直接运行即可。
代码如下：

% save jpg images as bin file for cpp
%
is_server = 1;

dataset = 'voc2012';  %'coco', 'voc2012'

if is_server
  if strcmp(dataset, 'voc2012')
    img_folder  = '/home/mmt/Desktop/my_dataset/VOC_aug/dataset/JPEGImages'
    save_folder = '/home/mmt/Desktop/my_dataset/VOC_aug/dataset/PPMImages';
  elseif strcmp(dataset, 'coco')
    img_folder  = '/rmt/data/coco/JPEGImages';
    save_folder = '/rmt/data/coco/PPMImages';
  end
else
  img_folder = '../img';
  save_folder = '../img_ppm';
end

if ~exist(save_folder, 'dir')
    mkdir(save_folder);
end

img_dir = dir(fullfile(img_folder, '*.jpg'));

for i = 1 : numel(img_dir)
    fprintf(1, 'processing %d (%d)...\n', i, numel(img_dir));
    img = imread(fullfile(img_folder, img_dir(i).name));

    img_fn = img_dir(i).name(1:end-4);
    save_fn = fullfile(save_folder, [img_fn, '.ppm']);

    imwrite(img, save_fn);   
end    

2）接下来，修改 run_densecrf.sh， 注意把 MODEL_NAME=deeplab_largeFOV修改了，原文件少了一个 p。

DATASET=voc2012  修改；
SAVE_DIR=/home/mmt/Desktop/deeplab_v2/${DATASET}/res/${FEATURE_NAME}/${MODEL_NAME}/${TEST_SET} 
修改；
CRF_DIR=/home/mmt/Desktop/deeplab_v2/deeplab-public-ver2/densecrf  修改；

if [ ${DATASET} == "voc2012" ]
then
    IMG_DIR_NAME=VOC_aug/dataset     修改；


FEATURE_DIR=/home/mmt/Desktop/deeplab_v2/${DATASET}/${FEATURE_NAME}/${MODEL_NAME}/${TEST_SET}/${FEATURE_TYPE}  修改；
同时把一些不需要的语句都注释掉，要不然容易出错，显示找不到文件。
修改后的文件如下：
#!/bin/bash 

###########################################
# You can either use this script to generate the DenseCRF post-processed results
# or use the densecrf_layer (wrapper) in Caffe
###########################################
DATASET=voc2012
LOAD_MAT_FILE=1

MODEL_NAME=deeplab_largeFOV

TEST_SET=val           #val, test

# the features  folder save the features computed via the model trained with the train set
# the features2 folder save the features computed via the model trained with the trainval set
FEATURE_NAME=features #features, features2
FEATURE_TYPE=fc8

# specify the parameters
MAX_ITER=10

Bi_W=4
Bi_X_STD=49
Bi_Y_STD=49
Bi_R_STD=5
Bi_G_STD=5 
Bi_B_STD=5

POS_W=3
POS_X_STD=3
POS_Y_STD=3


#######################################
# MODIFY THE PATY FOR YOUR SETTING
#######################################
SAVE_DIR=/home/mmt/Desktop/deeplab_v2/${DATASET}/res/${FEATURE_NAME}/${MODEL_NAME}/${TEST_SET}/${FEATURE_TYPE}/post_densecrf_W${Bi_W}_XStd${Bi_X_STD}_RStd${Bi_R_STD}_PosW${POS_W}_PosXStd${POS_X_STD}

echo "SAVE TO ${SAVE_DIR}"

CRF_DIR=/home/mmt/Desktop/deeplab_v2/deeplab-public-ver2/densecrf

#if [ ${DATASET} == "voc2012" ]
#then
    IMG_DIR_NAME=VOC_aug/dataset
#elif [ ${DATASET} == "coco" ]
#then
 #   IMG_DIR_NAME=coco
#elif [ ${DATASET} == "voc10_part" ]
#then
  #  IMG_DIR_NAME=pascal/VOCdevkit/VOC2012
#fi

# NOTE THAT the densecrf code only loads ppm images
IMG_DIR=/home/mmt/Desktop/my_dataset/${IMG_DIR_NAME}/PPMImages

#if [ ${LOAD_MAT_FILE} == 1 ]
#then
    # the features are saved in .mat format
    CRF_BIN=${CRF_DIR}/prog_refine_pascal_v4
    FEATURE_DIR=/home/mmt/Desktop/deeplab_v2/${DATASET}/${FEATURE_NAME}/${MODEL_NAME}/${TEST_SET}/${FEATURE_TYPE}
#else
    # the features are saved in .bin format (has called SaveMatAsBin.m in the densecrf/my_script)
   # CRF_BIN=${CRF_DIR}/prog_refine_pascal
   # FEATURE_DIR=/home/mmt/Desktop/deeplab_v2/${DATASET}/${FEATURE_NAME}/${MODEL_NAME}/${TEST_SET}/${FEATURE_TYPE}/bin
#fi

mkdir -p ${SAVE_DIR}

# run the program
${CRF_BIN} -id ${IMG_DIR} -fd ${FEATURE_DIR} -sd ${SAVE_DIR} -i ${MAX_ITER} -px ${POS_X_STD} -py ${POS_Y_STD} -pw ${POS_W} -bx ${Bi_X_STD} -by ${Bi_Y_STD} -br ${Bi_R_STD} -bg ${Bi_G_STD} -bb ${Bi_B_STD} -bw ${Bi_W}

进入文件路径，运行程序，结果如下图：
cd ~/Desktop/deeplab_v2/voc2012/
sh sh ./run_densecrf.sh

3）然后运行 /home/mmt/crf/deeplab-public-ver2/densecrf/my_script/GetDenseCRFResult.m把bin生成图片格式
注意修改文件路径（GetDenseCRFResult.m，SetupEnv在/deeplab_v2/deeplab-public-ver2/matlab/my_script中），
两个程序的代码如下：

GetDenseCRFResult.m
% compute the densecrf result (.bin) to png
%

addpath('/home/mmt/Desktop/deeplab_v2/deeplab-public-ver2/matlab/my_script');
SetupEnv;

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% You do not need to change values below
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
if is_server
  if learn_crf
    post_folder = sprintf('post_densecrf_W%d_XStd%d_RStd%d_PosW%d_PosXStd%d_ModelType%d_Epoch%d', bi_w, bi_x_std, bi_r_std, pos_w, pos_x_std, model_type, epoch);

    map_folder = fullfile('/home/mmt/Desktop/deeplab_v2', dataset, 'densecrf', 'res', feature_name, model_name, testset, feature_type, post_folder); 

    save_root_folder = fullfile('/home/mmt/Desktop/deeplab_v2', dataset, 'res', feature_name, model_name, testset, feature_type, post_folder); ;
  else
    post_folder = sprintf('post_densecrf_W%d_XStd%d_RStd%d_PosW%d_PosXStd%d', bi_w, bi_x_std, bi_r_std, pos_w, pos_x_std);

    map_folder = fullfile('/home/mmt/Desktop/deeplab_v2', dataset, 'res', feature_name, model_name, testset, feature_type, post_folder); 

    save_root_folder = map_folder;
  end
else 
  map_folder = '../result';
end

map_dir = dir(fullfile(map_folder, '*.bin'));


fprintf(1,' saving to %s\n', save_root_folder);

if strcmp(dataset, 'voc2012')
  seg_res_dir = [save_root_folder '/results/VOC2012/'];
elseif strcmp(dataset, 'coco')
  seg_res_dir = [save_root_folder, '/results/COCO2014/'];
else
  error('Wrong dataset!');
end

save_result_folder = fullfile(seg_res_dir, 'Segmentation', [id '_' testset '_cls']);

if ~exist(save_result_folder, 'dir')
    mkdir(save_result_folder);
end

for i = 1 : numel(map_dir)
    fprintf(1, 'processing %d (%d)...\n', i, numel(map_dir));
    map = LoadBinFile(fullfile(map_folder, map_dir(i).name), 'int16');

    img_fn = map_dir(i).name(1:end-4);
    imwrite(uint8(map), colormap, fullfile(save_result_folder, [img_fn, '.png']));
end

SetupEnv.m
% set up the environment variables
%

clear all; close all;
load('./pascal_seg_colormap.mat');

is_server       = 1;

crf_load_mat    = 1;   % the densecrf code load MAT files directly (no call SaveMatAsBin.m)
                       % used ONLY by DownSampleFeature.m
learn_crf       = 0;   % NOT USED. Set to 0

is_mat          = 1;   % the results to be evaluated are saved as mat (1) or png (0)
has_postprocess = 0;   % has done densecrf post processing (1) or not (0)
is_argmax       = 0;   % the output has been taken argmax already (e.g., coco dataset). 
                       % assume the argmax takes C-convention (i.e., start from 0)

debug           = 0;   % if debug, show some results

% vgg128_noup (not optimized well), aka DeepLab
% bi_w = 5, bi_x_std = 50, bi_r_std = 10

% vgg128_ms_pool3, aka DeepLab-MSc
% bi_w = 3, bi_x_std = 95, bi_r_std = 3

% vgg128_noup_pool3_cocomix, aka DeepLab-COCO
% bi_w = 5, bi_x_std = 67, bi_r_std = 3

%% these are used for the bounding box weak annotation experiments (i.e., to generate the Bbox-Seg)
% erode_gt (bbox)
% bi_w = 41, bi_x_std = 33, bi_r_std = 4

% erode_gt/bboxErode20
% bi_w = 45, bi_x_std = 37, bi_r_std = 3, pos_w = 15, pos_x_std = 3


%
% initial or default values for crf
%% 这几个参数要修改与run_densecrf.sh中的一致。
bi_w           = 4; 
bi_x_std       = 49;
bi_r_std       = 5;

pos_w          = 3;
pos_x_std      = 3;


%
dataset    = 'voc2012';  %'voc12', 'coco'  修改
trainset   = 'train_aug';      % not used
testset    = 'val';            %'val', 'test'

model_name = 'deeplab_largeFOV';  % 修改

feature_name = 'features';
feature_type = 'fc8'; % fc8 / crf

id           = 'comp6';

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% used for cross-validation
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
rng(10)

% downsampling files for cross-validation
down_sample_method = 2;      % 1: equally sample with "down_sample_rate", 2: randomly pick "num_sample" samples
down_sample_rate   = 8;
num_sample         = 100;    % number of samples used for cross-validation

% ranges for cross-validation
range_pos_w = [3];
range_pos_x_std = [3];

range_bi_w = [5];
range_bi_x_std = [49];
range_bi_r_std = [4 5];

4)至此，deeplabv2 程序已调试完。

总结，尝试过很多坑，终于除了结果。

---