CaffeOnACL/android/caffe_mobile.cpp at master · finley-/CaffeOnACL · GitHub

Repository navigation

History

executable file

273 lines (221 loc) · 8.43 KB

Raw

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

#include <algorithm>

#include <string>

#include <vector>

#include "boost/algorithm/string.hpp"

#include "caffe/caffe.hpp"

#include "caffe/layers/memory_data_layer.hpp"

#include "caffe_mobile.hpp"

#include <opencv2/core/core.hpp>

#include <opencv2/highgui/highgui.hpp>

#include <opencv2/imgproc/imgproc.hpp>

using std::clock;

using std::clock_t;

using std::string;

using std::vector;

using caffe::Blob;

using caffe::Caffe;

using caffe::Datum;

using caffe::Net;

using caffe::MemoryDataLayer;

namespace caffe {

template <typename T> vector<int> argmax(vector<T> const &values, int N) {

vector<size_t> indices(values.size());

std::iota(indices.begin(), indices.end(), static_cast<size_t>(0));

std::partial_sort(indices.begin(), indices.begin() + N, indices.end(),

[&](size_t a, size_t b) { return values[a] > values[b]; });

return vector<int>(indices.begin(), indices.begin() + N);

}

CaffeMobile *CaffeMobile::caffe_mobile_ = 0;

string CaffeMobile::model_path_ = "";

string CaffeMobile::weights_path_ = "";

CaffeMobile *CaffeMobile::Get() {

CHECK(caffe_mobile_);

return caffe_mobile_;

}

CaffeMobile *CaffeMobile::Get(const string &model_path,

const string &weights_path) {

if (!caffe_mobile_ || model_path != model_path_ ||

weights_path != weights_path_) {

caffe_mobile_ = new CaffeMobile(model_path, weights_path);

model_path_ = model_path;

weights_path_ = weights_path;

}

return caffe_mobile_;

}

CaffeMobile::CaffeMobile(const string &model_path, const string &weights_path) {

CHECK_GT(model_path.size(), 0) << "Need a model definition to score.";

CHECK_GT(weights_path.size(), 0) << "Need model weights to score.";

Caffe::set_mode(Caffe::CPU);

clock_t t_start = clock();

net_.reset(new Net<float>(model_path, caffe::TEST));

net_->CopyTrainedLayersFrom(weights_path);

clock_t t_end = clock();

LOG(INFO) << "Loading time: " << 1000.0 * (t_end - t_start) / CLOCKS_PER_SEC

<< " ms.";

CHECK_EQ(net_->num_inputs(), 1) << "Network should have exactly one input.";

CHECK_EQ(net_->num_outputs(), 1) << "Network should have exactly one output.";

Blob<float> *input_layer = net_->input_blobs()[0];

num_channels_ = input_layer->channels();

CHECK(num_channels_ == 3 || num_channels_ == 1)

<< "Input layer should have 1 or 3 channels.";

input_geometry_ = cv::Size(input_layer->width(), input_layer->height());

scale_ = 0.0;

}

CaffeMobile::~CaffeMobile() { net_.reset(); }

void CaffeMobile::SetMean(const vector<float> &mean_values) {

CHECK_EQ(mean_values.size(), num_channels_)

<< "Number of mean values doesn't match channels of input layer.";

cv::Scalar channel_mean(0);

double *ptr = &channel_mean[0];

for (int i = 0; i < num_channels_; ++i) {

ptr[i] = mean_values[i];

}

mean_ = cv::Mat(input_geometry_, (num_channels_ == 3 ? CV_32FC3 : CV_32FC1),

channel_mean);

}

void CaffeMobile::SetMean(const string &mean_file) {

BlobProto blob_proto;

ReadProtoFromBinaryFileOrDie(mean_file.c_str(), &blob_proto);

/* Convert from BlobProto to Blob<float> */

Blob<float> mean_blob;

mean_blob.FromProto(blob_proto);

CHECK_EQ(mean_blob.channels(), num_channels_)

<< "Number of channels of mean file doesn't match input layer.";

/* The format of the mean file is planar 32-bit float BGR or grayscale. */

std::vector<cv::Mat> channels;

float *data = mean_blob.mutable_cpu_data();

for (int i = 0; i < num_channels_; ++i) {

/* Extract an individual channel. */

cv::Mat channel(mean_blob.height(), mean_blob.width(), CV_32FC1, data);

channels.push_back(channel);

data += mean_blob.height() * mean_blob.width();

}

/* Merge the separate channels into a single image. */

cv::Mat mean;

cv::merge(channels, mean);

/* Compute the global mean pixel value and create a mean image

* filled with this value. */

cv::Scalar channel_mean = cv::mean(mean);

mean_ = cv::Mat(input_geometry_, mean.type(), channel_mean);

}

void CaffeMobile::SetScale(const float scale) {

CHECK_GT(scale, 0);

scale_ = scale;

}

void CaffeMobile::Preprocess(const cv::Mat &img,

std::vector<cv::Mat> *input_channels) {

/* Convert the input image to the input image format of the network. */

cv::Mat sample;

if (img.channels() == 3 && num_channels_ == 1)

cv::cvtColor(img, sample, cv::COLOR_BGR2GRAY);

else if (img.channels() == 4 && num_channels_ == 1)

cv::cvtColor(img, sample, cv::COLOR_BGRA2GRAY);

else if (img.channels() == 4 && num_channels_ == 3)

cv::cvtColor(img, sample, cv::COLOR_BGRA2BGR);

else if (img.channels() == 1 && num_channels_ == 3)

cv::cvtColor(img, sample, cv::COLOR_GRAY2BGR);

else

sample = img;

cv::Mat sample_resized;

if (sample.size() != input_geometry_)

cv::resize(sample, sample_resized, input_geometry_);

else

sample_resized = sample;

cv::Mat sample_float;

if (num_channels_ == 3)

sample_resized.convertTo(sample_float, CV_32FC3);

else

sample_resized.convertTo(sample_float, CV_32FC1);

cv::Mat sample_normalized;

if (!mean_.empty()) {

cv::subtract(sample_float, mean_, sample_normalized);

} else {

sample_normalized = sample_float;

}

if (scale_ > 0.0) {

sample_normalized *= scale_;

}

/* This operation will write the separate BGR planes directly to the

* input layer of the network because it is wrapped by the cv::Mat

* objects in input_channels. */

cv::split(sample_normalized, *input_channels);

CHECK(reinterpret_cast<float *>(input_channels->at(0).data) ==

net_->input_blobs()[0]->cpu_data())

<< "Input channels are not wrapping the input layer of the network.";

}

void CaffeMobile::WrapInputLayer(std::vector<cv::Mat> *input_channels) {

Blob<float> *input_layer = net_->input_blobs()[0];

int width = input_layer->width();

int height = input_layer->height();

float *input_data = input_layer->mutable_cpu_data();

for (int i = 0; i < input_layer->channels(); ++i) {

cv::Mat channel(height, width, CV_32FC1, input_data);

input_channels->push_back(channel);

input_data += width * height;

}

vector<float> CaffeMobile::Forward(const cv::Mat &img) {

CHECK(!img.empty()) << "img should not be empty";

Blob<float> *input_layer = net_->input_blobs()[0];

input_layer->Reshape(1, num_channels_, input_geometry_.height,

input_geometry_.width);

/* Forward dimension change to all layers. */

net_->Reshape();

vector<cv::Mat> input_channels;

WrapInputLayer(&input_channels);

Preprocess(img, &input_channels);

clock_t t_start = clock();

net_->Forward();

clock_t t_end = clock();

LOG(INFO) << "Forwarding time: " << 1000.0 * (t_end - t_start) / CLOCKS_PER_SEC

<< " ms.";

/* Copy the output layer to a std::vector */

Blob<float> *output_layer = net_->output_blobs()[0];

const float *begin = output_layer->cpu_data();

const float *end = begin + output_layer->channels();

return vector<float>(begin, end);

}

vector<float> CaffeMobile::GetConfidenceScore(const cv::Mat &img) {

return Forward(img);

}

vector<int> CaffeMobile::PredictTopK(const cv::Mat &img, int k) {

const vector<float> probs = Forward(img);

k = std::min<int>(std::max(k, 1), probs.size());

return argmax(probs, k);

}

vector<vector<float>>

CaffeMobile::ExtractFeatures(const cv::Mat &img,

const string &str_blob_names) {

Forward(img);

vector<std::string> blob_names;

boost::split(blob_names, str_blob_names, boost::is_any_of(","));

size_t num_features = blob_names.size();

for (size_t i = 0; i < num_features; i++) {

CHECK(net_->has_blob(blob_names[i])) << "Unknown feature blob name "

<< blob_names[i];

}

vector<vector<float>> features;

for (size_t i = 0; i < num_features; i++) {

const shared_ptr<Blob<float>> &feat = net_->blob_by_name(blob_names[i]);

features.push_back(

vector<float>(feat->cpu_data(), feat->cpu_data() + feat->count()));

}

return features;

}

} // namespace caffe

using caffe::CaffeMobile;

int main(int argc, char const *argv[]) {

string usage("usage: main <model> <weights> <mean_file> <img>");

if (argc < 5) {

std::cerr << usage << std::endl;

return 1;

}

CaffeMobile *caffe_mobile =

CaffeMobile::Get(string(argv[1]), string(argv[2]));

caffe_mobile->SetMean(string(argv[3]));

vector<int> top_3 = caffe_mobile->PredictTopK(cv::imread(string(argv[4]), -1), 3);

for (auto i : top_3) {

std::cout << i << std::endl;

}

return 0;

}

CaffeOnACL/android/caffe_mobile.cpp at master · finley-/CaffeOnACL · GitHub

Repository navigation

Expand file tree

Breadcrumbs

caffe_mobile.cpp

Latest commit

History

Breadcrumbs

caffe_mobile.cpp

File metadata and controls