【ビルド版】Raspberry Piで DeepDreamを動かしてキモイ絵をモリモリ量産 Caffe Deep Learning Framework

【インストール版】Raspberry Piで DeepDreamを動かしてキモイ絵をモリモリ量産 Caffe Deep Learning

pi@raspberrypi:~/pytorch $ uname -a
Linux raspberrypi 4.14.50-v7+ #1122 SMP Tue Jun 19 12:26:26 BST 2018 armv7l GNU/Linux

● Prerequisites
 BLAS via ATLAS, MKL, or OpenBLAS.
 Boost >= 1.55
 protobuf, glog, gflags, hdf5

● Pycaffe interfaces have their own natural needs.
 For Python Caffe: Python 2.7 or Python 3.3+, numpy (>= 1.7), boost-provided boost.python

● CPU-only Caffe
 for cold-brewed CPU-only Caffe uncomment the CPU_ONLY := 1 flag in Makefile.config to configure and build Caffe without CUDA. This is helpful for cloud or cluster deployment.

● BLAS
 OpenBLAS: free and open source; this optimized and parallel BLAS could require more effort to install, although it might offer a speedup.
 Install OpenBLAS
 Set BLAS := open in Makefile.config

● Python Caffe (optional)
 The main requirements are numpy and boost.python (provided by boost). pandas is useful too and needed for some examples.
 You can install the dependencies with
 pip install -r requirements.txt

● Compilation with Make
cp Makefile.config.example Makefile.config
# Adjust Makefile.config (for example, if using Anaconda Python, or if cuDNN is desired)
make all
make test
make runtest

● General dependencies
sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

● BLAS
 install ATLAS by
sudo apt-get install libatlas-base-dev

 install OpenBLAS by
sudo apt-get install libopenblas-dev

 MKL for better CPU performance.
※ MKLはラズパイでは関係有りません（Intel CPU専用）

● Python (optional)
 if you use the default Python you will need to
sudo apt-get install python-dev
 package to have the Python headers for building the pycaffe interface.

# glog
wget https://github.com/google/glog/archive/v0.3.3.tar.gz
tar zxvf v0.3.3.tar.gz
cd glog-0.3.3
./configure
make && make install

# gflags
wget https://github.com/schuhschuh/gflags/archive/master.zip
unzip master.zip
cd gflags-master
mkdir build && cd build
export CXXFLAGS="-fPIC" && cmake .. && make VERBOSE=1
make && make install

# lmdb
git clone https://github.com/LMDB/lmdb
cd lmdb/libraries/liblmdb
make && make install

Raspberry Piでメモリを馬鹿食いするアプリ用に不要なサービスを停止してフリーメモリを増やす方法

# ラズパイでお決まりの sudo apt-get update
sudo apt-get update

# BVLC/caffeの Gitからソースリスト一式を clone（ダウンロード）する
cd
git clone https://github.com/BVLC/caffe.git --depth 1
cd
cd caffe

# Makefile.configファイルを作成（参考例のファイルからコピー）する
cp Makefile.config.example Makefile.config
# Adjust Makefile.config (for example, if using Anaconda Python, or if cuDNN is desired)

# Makefile.configファイルをラズパイ環境に合わせて編集する
nano Makefile.config

# ラズパイ環境は演算処理に CPUしか使えないので CPU_ONLY := 1の行のコメント文字＃を取って有効にする
# CPU-only switch (uncomment to build without GPU support).
CPU_ONLY := 1

# BLAS（Basic Linear Algebra Subprograms、ベクトルと行列に関する数値演算ライブラリの名称）
# 今回は BLASライブラリとして ATLAS実装（Automatically Tuned Linear Algebra Software）を使用します
# BLAS choice:
# atlas for ATLAS (default)
# mkl for MKL
# open for OpenBlas
BLAS := atlas
# BLAS := open

# hdf5関係で「ファイルが見つからないエラー」が出るのでパスを設定する
# https://support.hdfgroup.org/HDF5/
# Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/arm-linux-gnueabihf/hdf5/serial

# コマンドラインでサクッと Makefile.configファイルを変更する
sed -i 's/# CPU_ONLY := 1/CPU_ONLY := 1/' Makefile.config
sed -i 's/BLAS :=.*/BLAS := atlas/' Makefile.config
sed -i 's/^INCLUDE_DIRS :=.*/INCLUDE_DIRS := $(PYTHON_INCLUDE) \/usr\/local\/include \/usr\/include\/hdf5\/serial/' Makefile.config
sed -i 's/^LIBRARY_DIRS :=.*/LIBRARY_DIRS := $(PYTHON_LIB) \/usr\/local\/lib \/usr\/lib \/usr\/lib\/arm-linux-gnueabihf\/hdf5\/serial/' Makefile.config

# 編集内容を確認する
grep "BLAS " Makefile.config
grep "CPU" Makefile.config
grep "_DIRS :" Makefile.config

# 編集内容を確認する
pi@raspberrypi:~/caffe $ grep "BLAS " Makefile.config
# BLAS choice:
BLAS := atlas

pi@raspberrypi:~/caffe $ grep "CPU" Makefile.config
# CPU-only switch (uncomment to build without GPU support).
CPU_ONLY := 1

pi@raspberrypi:~/caffe $ grep "_DIRS :" Makefile.config
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/arm-linux-gnueabihf/hdf5/serial

sudo apt-get -y install protobuf-compiler libprotoc-dev libboost-dev libgflags-dev libgoogle-glog-dev

# Makefile.config, BLAS := atlas
sudo apt-get -y install libatlas-base-dev
# Makefile.config, BLAS := open
# sudo apt-get -y install libopenblas-dev
sudo apt-get -y install libhdf5-dev libleveldb-dev liblmdb-dev libopencv-dev libsnappy-dev
sudo apt-get -y install libboost-system-dev libboost-filesystem-dev libboost-thread-dev

# make pycaffe
sudo apt-get -y install libboost-python-dev

# python deepdreamer.py sample.jpg
sudo apt-get -y install python-skimage python-protobuf python-numpy python-pil

# Makefile.config, BLAS := atlas
sudo apt-get -y install libatlas-base-dev protobuf-compiler libprotoc-dev libboost-dev libgflags-dev libgoogle-glog-dev libhdf5-dev libleveldb-dev liblmdb-dev libopencv-dev libsnappy-dev libboost-system-dev libboost-filesystem-dev libboost-thread-dev libboost-python-dev python-skimage python-protobuf python-numpy python-pil

# Makefile.config, BLAS := open
# sudo apt-get -y install libopenblas-dev

# libturbojpegをインストールする
sudo apt-get -y install libturbojpeg0 libturbojpeg0-dev
sudo ldconfig

# libboost-regex-devをインストールする
sudo apt-get -y install libboost-regex-dev

# libopenblas-base libopenblas-devをインストールする
sudo apt-get -y install libopenblas-base libopenblas-dev

# make clean

# make all -j3で 3 coreを使ってビルド速度を高速化（4 coreはあまり意味が無い）
# make all -j3で 3 coreを使う場合は上記の「メモリ使用量を減らす」を行なう事（ビルド中にメモリ不足でエラーになる）
# 「メモリ使用量を減らす」を行なわない場合は make all -j2で 2 coreビルドが無難。
date >BUILD_START
time make all -j3
date >BUILD_END
cat BUILD_START
cat BUILD_END

# make testと make runtest
make test -j3
make runtest -j3

# Python用の Caffeの作成と distributeでビルド内容を反映
make pycaffe
make distribute

# ここまで成功すれば Caffeのビルドは完了

# 環境変数 CAFFE_HOMEに PATHを設定する
# export CAFFE_HOME=/home/pi/caffe
export CAFFE_HOME=$(pwd)

echo ${CAFFE_HOME}

# パスにビルドで作成した Caffeを追加する
export PYTHONPATH=${CAFFE_HOME}/python:$PYTHONPATH
export LD_LIBRARY_PATH=${CAFFE_HOME}/distribute/lib:$LD_LIBRARY_PATH

pi@raspberrypi:~/caffe $ pwd
/home/pi/caffe

pi@raspberrypi:~/caffe $ export CAFFE_HOME=/home/pi/caffe

pi@raspberrypi:~/caffe $ echo $CAFFE_HOME
/home/pi/caffe

# 1コアでビルドで 26分程度
# real 25分41秒（1541sec）で、4コア時の real 584secより 2.6倍遅いです。
# real    25m41.666s
# user    24m58.530s
# sys     0m42.284s
# Sat 11 Aug 18:00:40 UTC 2018
# Sat 11 Aug 18:26:21 UTC 2018

# 2コアでビルドで 15分程度
# real    14m55.817s
# user    28m56.790s
# sys     0m50.334s
# Sat 11 Aug 18:30:49 UTC 2018
# Sat 11 Aug 18:45:45 UTC 2018

# 3コアでビルドで 11分程度
# real    11m18.701s
# user    32m44.597s
# sys     0m56.491s
# Sat 11 Aug 19:00:28 UTC 2018
# Sat 11 Aug 19:11:47 UTC 2018

# 4コアでビルドで 10分程度
# real    9m44.505s
# user    37m10.403s
# sys     1m9.108s
# Sat 11 Aug 17:49:48 UTC 2018
# Sat 11 Aug 17:59:32 UTC 2018

pi@raspberrypi:~/caffe $ make all -j4
PROTOC src/caffe/proto/caffe.proto
CXX .build_release/src/caffe/proto/caffe.pb.cc
CXX src/caffe/solvers/nesterov_solver.cpp
CXX src/caffe/solvers/adagrad_solver.cpp
...
AR -o .build_release/lib/libcaffe.a
LD -o .build_release/lib/libcaffe.so.1.0.0
CXX/LD -o .build_release/tools/upgrade_net_proto_binary.bin
CXX/LD -o .build_release/tools/upgrade_net_proto_text.bin
CXX/LD -o .build_release/tools/upgrade_solver_proto_text.bin
CXX/LD -o .build_release/tools/convert_imageset.bin
CXX/LD -o .build_release/tools/compute_image_mean.bin
CXX/LD -o .build_release/tools/extract_features.bin
CXX/LD -o .build_release/tools/caffe.bin
CXX/LD -o .build_release/examples/siamese/convert_mnist_siamese_data.bin
CXX/LD -o .build_release/examples/cpp_classification/classification.bin
CXX/LD -o .build_release/examples/cifar10/convert_cifar_data.bin
CXX/LD -o .build_release/examples/mnist/convert_mnist_data.bin

pi@raspberrypi:~/caffe $ ./.build_release/tools/caffe -version
caffe version 1.0.0

pi@raspberrypi:~/caffe $ ./.build_release/tools/caffe
caffe: command line brew
usage: caffe <command> <args>

commands:
  train           train or finetune a model
  test            score a model
  device_query    show GPU diagnostic information
  time            benchmark model execution time

  Flags from tools/caffe.cpp:
    -gpu (Optional; run in GPU mode on given device IDs separated by ','.Use
      '-gpu all' to run on all available GPUs. The effective training batch
      size is multiplied by the number of devices.) type: string default: ""
    -iterations (The number of iterations to run.) type: int32 default: 50
    -level (Optional; network level.) type: int32 default: 0
    -model (The model definition protocol buffer text file.) type: string
      default: ""
    -phase (Optional; network phase (TRAIN or TEST). Only used for 'time'.)
      type: string default: ""
    -sighup_effect (Optional; action to take when a SIGHUP signal is received:
      snapshot, stop or none.) type: string default: "snapshot"
    -sigint_effect (Optional; action to take when a SIGINT signal is received:
      snapshot, stop or none.) type: string default: "stop"
    -snapshot (Optional; the snapshot solver state to resume training.)
      type: string default: ""
    -solver (The solver definition protocol buffer text file.) type: string
      default: ""
    -stage (Optional; network stages (not to be confused with phase), separated
      by ','.) type: string default: ""
    -weights (Optional; the pretrained weights to initialize finetuning,
      separated by ','. Cannot be set simultaneously with snapshot.)
      type: string default: ""

real    18m44.714s
user    52m4.221s
sys     1m43.358s

pi@raspberrypi:~/caffe $ make test -j4
CXX src/caffe/test/test_maxpool_dropout_layers.cpp
CXX src/caffe/test/test_im2col_layer.cpp
CXX src/caffe/test/test_tanh_layer.cpp
CXX src/caffe/test/test_data_transformer.cpp
...
LD .build_release/src/caffe/test/test_contrastive_loss_layer.o
LD .build_release/src/caffe/test/test_pooling_layer.o
LD .build_release/src/caffe/test/test_reshape_layer.o
LD .build_release/src/caffe/test/test_solver.o
LD .build_release/src/caffe/test/test_convolution_layer.o
LD .build_release/src/caffe/test/test_threshold_layer.o
LD .build_release/src/caffe/test/test_lrn_layer.o
LD .build_release/src/caffe/test/test_slice_layer.o
LD .build_release/src/caffe/test/test_scale_layer.o

real    5m11.137s
user    4m48.411s
sys     0m2.134s

pi@raspberrypi:~/caffe $ make runtest -j4
...
[ RUN      ] UniformFillerTest/1.TestFill1D
[       OK ] UniformFillerTest/1.TestFill1D (0 ms)
[----------] 4 tests from UniformFillerTest/1 (1 ms total)

[----------] Global test environment tear-down
[==========] 1158 tests from 152 test cases ran. (302981 ms total)
[  PASSED  ] 1158 tests.

pi@raspberrypi:~/caffe $ make pycaffe
CXX/LD -o python/caffe/_caffe.so python/caffe/_caffe.cpp
touch python/caffe/proto/__init__.py
PROTOC (python) src/caffe/proto/caffe.proto

pi@raspberrypi:~/caffe $ make distribute
# add proto
cp -r src/caffe/proto distribute/
# add include
cp -r include distribute/
mkdir -p distribute/include/caffe/proto
cp .build_release/src/caffe/proto/caffe.pb.h distribute/include/caffe/proto
# add tool and example binaries
cp .build_release/tools/upgrade_solver_proto_text.bin .build_release/tools/upgrade_net_proto_binary.bin .build_release/tools/upgrade_net_proto_text.bin .build_release/tools/convert_imageset.bin .build_release/tools/extract_features.bin .build_release/tools/compute_image_mean.bin .build_release/tools/caffe.bin distribute/bin
cp .build_release/examples/siamese/convert_mnist_siamese_data.bin .build_release/examples/cpp_classification/classification.bin .build_release/examples/cifar10/convert_cifar_data.bin .build_release/examples/mnist/convert_mnist_data.bin distribute/bin
# add libraries
cp .build_release/lib/libcaffe.a distribute/lib
install -m 644 .build_release/lib/libcaffe.so.1.0.0 distribute/lib
cd distribute/lib; rm -f libcaffe.so;   ln -s libcaffe.so.1.0.0 libcaffe.so
# add python - it's not the standard way, indeed...
cp -r python distribute/

pi@raspberrypi:~/caffe $ ls -l distribute/
total 20
drwxr-xr-x 2 pi pi 4096 Aug  8 11:08 bin
drwxr-xr-x 3 pi pi 4096 Aug  8 11:08 include
drwxr-xr-x 2 pi pi 4096 Aug  8 11:08 lib
drwxr-xr-x 2 pi pi 4096 Aug  8 11:08 proto
drwxr-xr-x 3 pi pi 4096 Aug  8 11:08 python

pi@raspberrypi:~/caffe $ ls -l distribute/bin/
total 456
-rwxr-xr-x 1 pi pi 113004 Aug  8 11:08 caffe.bin
-rwxr-xr-x 1 pi pi  61648 Aug  8 11:08 classification.bin
-rwxr-xr-x 1 pi pi  21076 Aug  8 11:08 compute_image_mean.bin
-rwxr-xr-x 1 pi pi  30064 Aug  8 11:08 convert_cifar_data.bin
-rwxr-xr-x 1 pi pi  38368 Aug  8 11:08 convert_imageset.bin
-rwxr-xr-x 1 pi pi  26464 Aug  8 11:08 convert_mnist_data.bin
-rwxr-xr-x 1 pi pi  26248 Aug  8 11:08 convert_mnist_siamese_data.bin
-rwxr-xr-x 1 pi pi  80584 Aug  8 11:08 extract_features.bin
-rwxr-xr-x 1 pi pi  14372 Aug  8 11:08 upgrade_net_proto_binary.bin
-rwxr-xr-x 1 pi pi  14364 Aug  8 11:08 upgrade_net_proto_text.bin
-rwxr-xr-x 1 pi pi  14392 Aug  8 11:08 upgrade_solver_proto_text.bin

pi@raspberrypi:~/caffe $ ls -l distribute/lib/
total 17524
-rw-r--r-- 1 pi pi 13906940 Aug  8 11:08 libcaffe.a
lrwxrwxrwx 1 pi pi       17 Aug  8 11:08 libcaffe.so -> libcaffe.so.1.0.0
-rw-r--r-- 1 pi pi  4031784 Aug  8 11:08 libcaffe.so.1.0.0

pi@raspberrypi:~/caffe $ ls -l distribute/python/
total 36
drwxr-xr-x 5 pi pi 4096 Aug  8 11:08 caffe
-rwxr-xr-x 1 pi pi 4262 Aug  8 11:08 classify.py
-rw-r--r-- 1 pi pi 1523 Aug  8 11:08 CMakeLists.txt
-rwxr-xr-x 1 pi pi 5734 Aug  8 11:08 detect.py
-rwxr-xr-x 1 pi pi 2254 Aug  8 11:08 draw_net.py
-rw-r--r-- 1 pi pi  261 Aug  8 11:08 requirements.txt
-rw-r--r-- 1 pi pi 3145 Aug  8 11:08 train.py

Raspberry Piに Jupyter Notebookをインストールして拡張子 ipynb形式の IPythonを動かす

# Python版の Google Deep Dream kesara/deepdreamerをダウンロードする
cd
git clone https://github.com/kesara/deepdreamer
cd deepdreamer

# 学習データ bvlc_googlenet.caffemodelをダウンロードする
wget http://dl.caffe.berkeleyvision.org/bvlc_googlenet.caffemodel

# deploy.prototxtをダウンロードする
# https://github.com/BVLC/caffe/tree/master/models/bvlc_googlenet
wget https://raw.githubusercontent.com/BVLC/caffe/master/models/bvlc_googlenet/deploy.prototxt

# deploy.prototxtの一番最後に force_backward: trueの行を追加する
# nano deploy.prototxt
# force_backward: true

# コマンドラインでサクッと force_backward: trueの行を追加する
echo "force_backward: true" >> deploy.prototxt

# tailコマンドで最後に追加された事を確認する
tail deploy.prototxt

pi@raspberrypi:~/deepdreamer $ python3 deepdreamer.py sample.jpg

Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 11, in <module>
    from caffe import Classifier, set_device, set_mode_gpu
  File "/home/pi/DeepDream/caffe/python/caffe/__init__.py", line 1, in <module>
    from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver, NCCL, Timer
  File "/home/pi/DeepDream/caffe/python/caffe/pycaffe.py", line 13, in <module>
    from ._caffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, \
ImportError: dynamic module does not define module export function (PyInit__caffe)

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py sample.jpg

Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 12, in <module>
    from deepdreamer.images2gif import writeGif
ImportError: No module named images2gif

nano ./deepdreamer/deepdreamer.py

● 12行目をコメントにする。
# from deepdreamer.images2gif import writeGif

● 196～198行目をコメントにする。
#         writeGif(
#             "{}.gif".format(img_path), frames, duration=duration,
#             repeat=loop)

# コマンドラインでサクッと deepdreamer/deepdreamer.pyを編集する
# 196～198行目を削除する
sed '196,198d' -i ./deepdreamer/deepdreamer.py
# 12行目を削除する
sed '12d' -i ./deepdreamer/deepdreamer.py

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py sample.jpg

cd
cd deepdreamer
wget https://github.com/google/deepdream/raw/master/flowers.jpg

python deepdreamer.py flowers.jpg

# Copy File
cp flowers.jpg flowers_3a.jpg
cp flowers.jpg flowers_3b.jpg

cp flowers.jpg flowers_4a.jpg
cp flowers.jpg flowers_4c.jpg
cp flowers.jpg flowers_4e.jpg

cp flowers.jpg flowers_5a.jpg
cp flowers.jpg flowers_5b.jpg

# 3
python deepdreamer.py flowers_3a.jpg --dreams 1 --itern 15 --layers inception_3a/output

python deepdreamer.py flowers_3b.jpg --dreams 1 --itern 15 --layers inception_3b/output

# 4
python deepdreamer.py flowers_4a.jpg --dreams 1 --itern 15 --layers inception_4a/output

python deepdreamer.py flowers_4c.jpg --dreams 1 --itern 15 --layers inception_4c/output

python deepdreamer.py flowers_4e.jpg --dreams 1 --itern 15 --layers inception_4e/output

# 5
python deepdreamer.py flowers_5a.jpg --dreams 1 --itern 15 --layers inception_5a/output

python deepdreamer.py flowers_5b.jpg --dreams 1 --itern 15 --layers inception_5b/output

pi@raspberrypi:~ $ export PYTHONPATH=/home/pi/DeepDream/caffe/python:$PYTHONPATH

pi@raspberrypi:~ $ export LD_LIBRARY_PATH=/home/pi/DeepDream/caffe/distribute/lib:$LD_LIBRARY_PATH

pi@raspberrypi:~/deepdreamer $ ls -l /home/pi/caffe/distribute/lib
total 17524
-rw-r--r-- 1 pi pi 13906940 Aug  4 12:50 libcaffe.a
lrwxrwxrwx 1 pi pi       17 Aug  4 12:50 libcaffe.so -> libcaffe.so.1.0.0
-rw-r--r-- 1 pi pi  4031784 Aug  4 12:50 libcaffe.so.1.0.0

pi@raspberrypi:~/deepdreamer $ ls -l /home/pi/caffe/python/
total 36
drwxr-xr-x 6 pi pi 4096 Aug  4 15:38 caffe
-rwxr-xr-x 1 pi pi 4262 Aug  4 08:25 classify.py
-rw-r--r-- 1 pi pi 1523 Aug  4 08:25 CMakeLists.txt
-rwxr-xr-x 1 pi pi 5734 Aug  4 08:25 detect.py
-rwxr-xr-x 1 pi pi 2254 Aug  4 08:25 draw_net.py
-rw-r--r-- 1 pi pi  261 Aug  4 08:25 requirements.txt
-rw-r--r-- 1 pi pi 3145 Aug  4 08:25 train.py

nano Makefile.config
BLAS := open

# Makefile.config, BLAS := open
sudo apt-get -y install libopenblas-dev

# 環境変数で OpenBLASライブラリを 4コアで動かす指定をする
export OPENBLAS_NUM_THREADS=4

# Setting the number of threads using environment variables
# The priorities are OPENBLAS_NUM_THREADS > GOTO_NUM_THREADS > OMP_NUM_THREADS .
# https://github.com/xianyi/OpenBLAS
export OPENBLAS_NUM_THREADS=4
export GOTO_NUM_THREADS=${OPENBLAS_NUM_THREADS}
export OMP_NUM_THREADS=${OPENBLAS_NUM_THREADS}

Dreaming...
/usr/lib/python2.7/dist-packages/scipy/ndimage/interpolation.py:568: UserWarning: From scipy 0.13.0, the output shape of zoom() is calculated with round() instead of int() - for these inputs the size of the returned array has changed.
  "the returned array has changed.", UserWarning)
Segmentation fault

export OPENBLAS_NUM_THREADS=2
Segmentation fault
マルチコアのマルチスレッドで動きません！

vcgencmd get_throttled

pi@raspberrypi:~/deepdreamer $ vcgencmd get_throttled
throttled=0x50000
Segmentation fault
マルチコアのマルチスレッドで動きません！

export OPENBLAS_NUM_THREADS=1
1を指定して 1コアで動かせば安定して動きました（意味ねー）

diff -ruN common_linux.h.orig common_linux.h
--- common_linux.h.orig	2012-04-23 11:27:55.000000000 +0800
+++ common_linux.h	2012-05-08 23:43:00.000000000 +0800
@@ -77,7 +77,7 @@
 #else
 //Fixed randomly SEGFAULT when nodemask==NULL with above Linux 2.6.34
 //	unsigned long null_nodemask=0;
-	return syscall(SYS_mbind, addr, len, mode, nodemask, maxnode, flags);
+	return 0;
 #endif
 }

pi@raspberrypi:~/caffe $ make all
PROTOC src/caffe/proto/caffe.proto
make: protoc: Command not found
Makefile:638: recipe for target '.build_release/src/caffe/proto/caffe.pb.cc' failed
make: *** [.build_release/src/caffe/proto/caffe.pb.cc] Error 127

# Solution
sudo apt-get -y install protobuf-compiler
sudo apt-get -y install libprotoc-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/solvers/adagrad_solver.cpp
In file included from ./include/caffe/sgd_solvers.hpp:7:0,
                 from src/caffe/solvers/adagrad_solver.cpp:3:
./include/caffe/solver.hpp:3:30: fatal error: boost/function.hpp: No such file or directory
 #include <boost/function.hpp>
                              ^
compilation terminated.

# Solution
apt-cache search boost
sudo apt-get -y install libboost-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/solvers/adam_solver.cpp
In file included from ./include/caffe/blob.hpp:8:0,
                 from ./include/caffe/net.hpp:10,
                 from ./include/caffe/solver.hpp:7,
                 from ./include/caffe/sgd_solvers.hpp:7,
                 from src/caffe/solvers/nesterov_solver.cpp:3:
./include/caffe/common.hpp:5:27: fatal error: gflags/gflags.h: No such file or directory
 #include <gflags/gflags.h>
                           ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/solvers/nesterov_solver.o' failed
make: *** [.build_release/src/caffe/solvers/nesterov_solver.o] Error 1

# Solution
sudo apt-get -y install libgflags-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/solvers/nesterov_solver.cpp
In file included from ./include/caffe/blob.hpp:8:0,
                 from ./include/caffe/net.hpp:10,
                 from ./include/caffe/solver.hpp:7,
                 from ./include/caffe/sgd_solvers.hpp:7,
                 from src/caffe/solvers/nesterov_solver.cpp:3:
./include/caffe/common.hpp:6:26: fatal error: glog/logging.h: No such file or directory
 #include <glog/logging.h>
                          ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/solvers/nesterov_solver.o' failed
make: *** [.build_release/src/caffe/solvers/nesterov_solver.o] Error 1

# Solution
sudo apt-get -y install libgoogle-glog-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/solvers/nesterov_solver.cpp
In file included from ./include/caffe/util/math_functions.hpp:11:0,
                 from ./include/caffe/layer.hpp:12,
                 from ./include/caffe/net.hpp:12,
                 from ./include/caffe/solver.hpp:7,
                 from ./include/caffe/sgd_solvers.hpp:7,
                 from src/caffe/solvers/nesterov_solver.cpp:3:
./include/caffe/util/mkl_alternate.hpp:14:19: fatal error: cblas.h: No such file or directory
 #include <cblas.h>
                   ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/solvers/nesterov_solver.o' failed
make: *** [.build_release/src/caffe/solvers/nesterov_solver.o] Error 1

# Solution
# BLAS 数値計算ライブラリ ATLASをインストール
# Makefile.config, BLAS := atlas
sudo apt-get -y install libatlas-base-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/solvers/sgd_solver.cpp
In file included from src/caffe/solvers/sgd_solver.cpp:5:0:
./include/caffe/util/hdf5.hpp:6:18: fatal error: hdf5.h: No such file or directory
 #include "hdf5.h"
                  ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/solvers/sgd_solver.o' failed
make: *** [.build_release/src/caffe/solvers/sgd_solver.o] Error 1

# Solution
# NG
# sudo apt-get install libhdf5-cpp-11
# E: Unable to locate package libhdf5-cpp-11

# OK
apt-cache search libhdf5
sudo apt-get -y install libhdf5-dev

# Solution 2
Makefile.config
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/arm-linux-gnueabihf/hdf5/serial

pi@raspberrypi:~ $ ls -l /usr/include/hdf5/serial/hdf5.h
-rw-r--r-- 1 root root 2719 Aug 19  2017 /usr/include/hdf5/serial/hdf5.h

pi@raspberrypi:~ $ ls -l /usr/lib/arm-linux-gnueabihf/hdf5/serial/
total 0
lrwxrwxrwx 1 root root 31 Aug 19  2017 include -> ../../../../include/hdf5/serial
lrwxrwxrwx 1 root root  1 Aug 19  2017 lib -> .
lrwxrwxrwx 1 root root 22 Aug 19  2017 libhdf5.a -> ../../libhdf5_serial.a
lrwxrwxrwx 1 root root 19 Aug 19  2017 libhdf5_cpp.a -> ../../libhdf5_cpp.a
lrwxrwxrwx 1 root root 20 Aug 19  2017 libhdf5_cpp.so -> ../../libhdf5_cpp.so
lrwxrwxrwx 1 root root 30 Aug 19  2017 libhdf5_fortran.a -> ../../libhdf5_serial_fortran.a

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/util/db_leveldb.cpp
In file included from src/caffe/util/db_leveldb.cpp:2:0:
./include/caffe/util/db_leveldb.hpp:7:24: fatal error: leveldb/db.h: No such file or directory
 #include "leveldb/db.h"
                        ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/util/db_leveldb.o' failed
make: *** [.build_release/src/caffe/util/db_leveldb.o] Error 1

# Solution
sudo apt-get -y install libleveldb-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/util/db_lmdb.cpp
In file included from src/caffe/util/db_lmdb.cpp:2:0:
./include/caffe/util/db_lmdb.hpp:8:18: fatal error: lmdb.h: No such file or directory
 #include "lmdb.h"
                  ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/util/db_lmdb.o' failed
make: *** [.build_release/src/caffe/util/db_lmdb.o] Error 1

# Solution
sudo apt-get -y install liblmdb-dev

pi@raspberrypi:~/caffe $ make all
...
CXX src/caffe/util/io.cpp
src/caffe/util/io.cpp:6:33: fatal error: opencv2/core/core.hpp: No such file or directory
 #include <opencv2/core/core.hpp>
                                 ^
compilation terminated.
Makefile:581: recipe for target '.build_release/src/caffe/util/io.o' failed
make: *** [.build_release/src/caffe/util/io.o] Error 1

# Solution
sudo apt-get -y install libopencv-dev

pi@raspberrypi:~/caffe $ make all
...
AR -o .build_release/lib/libcaffe.a
LD -o .build_release/lib/libcaffe.so.1.0.0
/usr/bin/ld: cannot find -lsnappy
collect2: error: ld returned 1 exit status
Makefile:572: recipe for target '.build_release/lib/libcaffe.so.1.0.0' failed
make: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1

# Solution
sudo apt-get -y install libsnappy-dev
# Setting up libsnappy-dev:armhf (1.1.3-3) ...

pi@raspberrypi:~/caffe $ make all
...
AR -o .build_release/lib/libcaffe.a
LD -o .build_release/lib/libcaffe.so.1.0.0
/usr/bin/ld: cannot find -lboost_system
/usr/bin/ld: cannot find -lboost_filesystem
/usr/bin/ld: cannot find -lboost_thread
collect2: error: ld returned 1 exit status
Makefile:572: recipe for target '.build_release/lib/libcaffe.so.1.0.0' failed
make: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1

# Solution
sudo apt-get -y install libboost-system-dev libboost-filesystem-dev libboost-thread-dev
# Setting up libboost-system1.62-dev:armhf (1.62.0+dfsg-4) ...
# Setting up libboost-filesystem-dev:armhf (1.62.0.1+b4) ...
# Setting up libboost-thread1.62-dev:armhf (1.62.0+dfsg-4) ...

pi@raspberrypi:~/caffe $ make pycaffe
CXX/LD -o python/caffe/_caffe.so python/caffe/_caffe.cpp
/usr/bin/ld: cannot find -lboost_python
collect2: error: ld returned 1 exit status
Makefile:507: recipe for target 'python/caffe/_caffe.so' failed
make: *** [python/caffe/_caffe.so] Error 1

# Solution
# NG sudo apt-get -y install libboost_python
# E: Unable to locate package libboost_python

apt-cache search libboost | grep python
sudo apt-get -y install libboost-python-dev
# Setting up libboost-python-dev (1.62.0.1+b4) ...

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 11, in <module>
    from caffe import Classifier, set_device, set_mode_gpu
ImportError: No module named caffe

# Solution
export PYTHONPATH=/home/pi/caffe/python:$PYTHONPATH
export LD_LIBRARY_PATH=/home/pi/caffe/distribute/lib:$LD_LIBRARY_PATH

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 11, in <module>
    from caffe import Classifier, set_device, set_mode_gpu
  File "/home/pi/caffe/python/caffe/__init__.py", line 1, in <module>
    from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver, NCCL, Timer
  File "/home/pi/caffe/python/caffe/pycaffe.py", line 15, in <module>
    import caffe.io
  File "/home/pi/caffe/python/caffe/io.py", line 2, in <module>
    import skimage.io
ImportError: No module named skimage.io

# Solution
sudo apt-get -y install python-skimage

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 11, in <module>
    from caffe import Classifier, set_device, set_mode_gpu
  File "/home/pi/caffe/python/caffe/__init__.py", line 1, in <module>
    from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver, NCCL, Timer
  File "/home/pi/caffe/python/caffe/pycaffe.py", line 15, in <module>
    import caffe.io
  File "/home/pi/caffe/python/caffe/io.py", line 8, in <module>
    from caffe.proto import caffe_pb2
  File "/home/pi/caffe/python/caffe/proto/caffe_pb2.py", line 6, in <module>
    from google.protobuf.internal import enum_type_wrapper
ImportError: No module named google.protobuf.internal

# Solution
sudo apt-get -y install python-protobuf

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/user/deepdreamer/deepdreamer/deepdreamer.py", line 10, in <module>
    import numpy as np
ImportError: No module named numpy

pi@raspberrypi:~ $ apt-cache search numpy | grep python-numpy
python-numpy - Numerical Python adds a fast array facility to the Python language
python-numpy-dbg - Fast array facility to the Python language (debug extension)
python-numpy-doc - NumPy documentation
python-numpydoc - Sphinx extension to support docstrings in Numpy format -- Python 2

# Solution
sudo apt-get -y install python-numpy

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 12, in <module>
    from deepdreamer.images2gif import writeGif
ImportError: No module named images2gif

# Solution
# images2gif import writeGifの gif動画書き出し機能を無効にする。
# コマンドラインでサクッと deepdreamer/deepdreamer.pyを編集する
sed '196,198d' -i ./deepdreamer/deepdreamer.py
sed '12d' -i ./deepdreamer/deepdreamer.py

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/user/deepdreamer/deepdreamer/deepdreamer.py", line 13, in <module>
    from PIL.Image import fromarray as img_fromarray, open as img_open
ImportError: No module named PIL.Image

# Solution
# pip install pillow
# The program 'pip' is currently not installed. You can install it by typing:
# sudo apt install python-pip
# pip install pillow
sudo apt-get -y install python-pil

pi@raspberrypi:~ $ apt-cache search pillow
python-pil - Python Imaging Library (Pillow fork)
python-pil.imagetk - Python Imaging Library - ImageTk Module (Pillow fork)
python-pyocr - Python wrapper for OCR engines (Python 2)
python-sane - Python Imaging Library - SANE interface (Pillow fork)
python-willow - Python image library combining Pillow, Wand and OpenCV
python-willow-doc - Python image library (documentation)
python3-pil - Python Imaging Library (Python3)
python3-pyocr - Python wrapper for OCR engines (Python 3)

pi@raspberrypi:~/deepdreamer $ python3 deepdreamer.py flowers.jpg
Traceback (most recent call last):
  File "deepdreamer.py", line 9, in <module>
    from deepdreamer.deepdreamer import deepdream, deepdream_video, list_layers
  File "/home/pi/deepdreamer/deepdreamer/deepdreamer.py", line 11, in <module>
    from caffe import Classifier, set_device, set_mode_gpu
  File "/home/pi/caffe/python/caffe/__init__.py", line 1, in <module>
    from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver, NCCL, Timer
  File "/home/pi/caffe/python/caffe/pycaffe.py", line 13, in <module>
    from ._caffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, \
ImportError: dynamic module does not define module export function (PyInit__caffe)

pi@raspberrypi:~/deepdreamer $ python deepdreamer.py sample.jpg

...

I0806 11:33:55.667161   583 layer_factory.hpp:77] Creating layer data
I0806 11:33:55.667289   583 net.cpp:84] Creating Layer data
I0806 11:33:55.667322   583 net.cpp:380] data -> data
I0806 11:33:55.667415   583 net.cpp:122] Setting up data
I0806 11:33:55.667438   583 net.cpp:129] Top shape: 10 3 224 224 (1505280)
I0806 11:33:55.667469   583 net.cpp:137] Memory required for data: 6021120
I0806 11:33:55.667491   583 layer_factory.hpp:77] Creating layer conv1/7x7_s2

...

I0806 11:34:01.427542   583 net.cpp:744] Ignoring source layer loss2/classifier
I0806 11:34:01.427561   583 net.cpp:744] Ignoring source layer loss2/loss
I0806 11:34:01.445107   583 net.cpp:744] Ignoring source layer loss3/loss3
Dreaming...
/usr/lib/python2.7/dist-packages/scipy/ndimage/interpolation.py:568: UserWarning: From scipy 0.13.0, the output shape of zoom() is calculated with round() instead of int() - for these inputs the size of the returned array has changed.
  "the returned array has changed.", UserWarning)
Dream 0 saved.
/usr/lib/python2.7/dist-packages/scipy/ndimage/interpolation.py:430: UserWarning: The behaviour of affine_transform with a one-dimensional array supplied for the matrix parameter has changed in scipy 0.18.0.
  "The behaviour of affine_transform with a one-dimensional "
Dream 1 saved.
Dream 2 saved.

...

cd
git clone https://github.com/xianyi/OpenBLAS.git
cd OpenBLAS

# SEGFAULT with multi-threading patch to disable mbind using
patch < segfaults.patch

# quick build 32bit
make -j 2 BINARY=32

# install
sudo make install

make[1]: Leaving directory '/home/pi/OpenBLAS/exports'

 OpenBLAS build complete. (BLAS CBLAS LAPACK LAPACKE)

  OS               ... Linux
  Architecture     ... arm
  BINARY           ... 32bit
  C compiler       ... GCC  (command line : cc)
  Fortran compiler ... GFORTRAN  (command line : gfortran)
  Library Name     ... libopenblas_armv7p-r0.3.3.dev.a (Multi threaded; Max num-threads is 4)

To install the library, you can run "make PREFIX=/path/to/your/installation install".

usage: deepdreamer.py [-h] [--zoom {true,false}] [--scale SCALE]
                      [--dreams DREAMS] [--itern ITERN] [--octaves OCTAVES]
                      [--octave-scale OCTAVE_SCALE] [--layers LAYERS]
                      [--clip {true,false}] [--gpuid GPUID]
                      [--network {bvlc_googlenet,googlenet_place205}]
                      [--gif {true,false}] [--reverse {true,false}]
                      [--duration DURATION] [--loop {true,false}]
                      [--framerate FRAMERATE] [--list-layers] [--video VIDEO]
                      [image]

positional arguments:
  image

optional arguments:
  -h, --help            show this help message and exit
  --gpuid GPUID         enable GPU with id GPUID (default: disabled)
  --zoom {true,false}   zoom dreams (default: true)
  --scale SCALE         scale coefficient for zoom (default: 0.05)
  --dreams DREAMS       number of images (default: 100)
  --itern ITERN         dream iterations (default: 10)
  --octaves OCTAVES     dream octaves (default: 4)
  --octave-scale OCTAVE_SCALE
                        dream octave scale (default: 1.4)
  --layers LAYERS       dream layers (default: inception_4c/output)
  --clip {true,false}   clip dreams (default: true)
  --network {bvlc_googlenet,googlenet_place205}
                        choose the network to use (default: bvlc_googlenet)
  --gif {true,false}    make a gif (default: false)
  --reverse {true,false}
                        make a reverse gif (default: false)
  --duration DURATION   gif frame duration in seconds (default: 0.1)
  --loop {true,false}   enable gif loop (default: false)
  --framerate FRAMERATE
                        framerate for video (default: 24)
  --list-layers         list layers
  --video VIDEO         video file

inception_3a/output
inception_3b/output

inception_4a/output
inception_4b/output
inception_4c/output *default
inception_4d/output
inception_4e/output

inception_5a/output
inception_5b/output

（抜粋）
...
layer {
  name: "inception_3b/output"
  type: "Concat"
  bottom: "inception_3b/1x1"
  bottom: "inception_3b/3x3"
  bottom: "inception_3b/5x5"
  bottom: "inception_3b/pool_proj"
  top: "inception_3b/output"
}
...
layer {
  name: "inception_4a/output"
  type: "Concat"
  bottom: "inception_4a/1x1"
  bottom: "inception_4a/3x3"
  bottom: "inception_4a/5x5"
  bottom: "inception_4a/pool_proj"
  top: "inception_4a/output"
}
...
layer {
  name: "inception_4e/output"
  type: "Concat"
  bottom: "inception_4e/1x1"
  bottom: "inception_4e/3x3"
  bottom: "inception_4e/5x5"
  bottom: "inception_4e/pool_proj"
  top: "inception_4e/output"
}
...
layer {
  name: "inception_5b/output"
  type: "Concat"
  bottom: "inception_5b/1x1"
  bottom: "inception_5b/3x3"
  bottom: "inception_5b/5x5"
  bottom: "inception_5b/pool_proj"
  top: "inception_5b/output"
}

cd
cd deepdreamer

wget http://places.csail.mit.edu/model/googlenet_places205.tar.gz

tar xvfz googlenet_places205.tar.gz

cp googlenet_places205/googlelet_places205_train_iter_2400000.caffemodel .
cp googlenet_places205/deploy_places205.protxt .

python deepdreamer.py flowers_4c.jpg --dreams 1 --itern 20 --layers inception_4c/output --network googlenet_place205
python deepdreamer.py flowers_3a.jpg --dreams 1 --itern 20 --layers inception_3a/output --network googlenet_place205
python deepdreamer.py flowers_5a.jpg --dreams 1 --itern 20 --layers inception_5a/output --network googlenet_place205
python deepdreamer.py flowers_3b.jpg --dreams 1 --itern 20 --layers inception_3b/output --network googlenet_place205
python deepdreamer.py flowers_5b.jpg --dreams 1 --itern 20 --layers inception_5b/output --network googlenet_place205
python deepdreamer.py flowers_4a.jpg --dreams 1 --itern 20 --layers inception_4a/output --network googlenet_place205
python deepdreamer.py flowers_4e.jpg --dreams 1 --itern 20 --layers inception_4e/output --network googlenet_place205

NVIDIA Jetson Nano 開発者キットを買ってみた。メモリ容量 4GB LPDDR4 RAM

Raspberry Piでメモリを馬鹿食いするアプリ用に不要なサービスを停止してフリーメモリを増やす方法

【成功版】最新版の Darknetに digitalbrain79版の Darknet with NNPACKの NNPACK処理を適用する

【成功版】Raspberry Piで NNPACK対応版の Darknet Neural Network Frameworkをビルドする方法

【成功版】Raspberry Piで Darknet Neural Network Frameworkをビルドする方法

【成功版】Raspberry Piに TensorFlow Deep Learning Frameworkをインストールする方法

Raspberry Piで TensorFlow Deep Learning Frameworkを自己ビルドする方法

Raspberry Piで Caffe Deep Learning Frameworkで物体認識を行なってみるテスト

【インストール版】Raspberry Piで DeepDreamを動かしてキモイ絵をモリモリ量産 Caffe Deep Learning

Raspberry Piで Caffe2 Deep Learning Frameworkをソースコードからビルドする方法

Orange Pi PC 2の 64bitのチカラで DeepDreamしてキモイ絵を高速でモリモリ量産してみるテスト

Raspberry Piに Jupyter Notebookをインストールして拡張子 ipynb形式の IPythonを動かす

Raspberry Piで Deep Learningフレームワーク Chainerをインストールしてみる

Raspberry Piで DeepBeliefSDKをビルドして画像認識フレームワークを動かす方法

Raspberry Piで Microsoftの ELLをビルドする方法

Raspberry Piで MXNet port of SSD Single Shot MultiBoxを動かして画像の物体検出をする方法

Raspberry Piで Apache MXNet Incubatingをビルドする方法

Raspberry Piで OpenCVの Haar Cascade Object Detectionでリアルタイムにカメラ映像の顔検出を行なってみる

Raspberry Piで NNPACKをビルドする方法

Raspberry Pi 3の Linuxコンソール上で使用する各種コマンドまとめ