Skip to content

P
Paddle

迷途归一 / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 0
Fork 0
P
Paddle

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
查找文件 普通视图历史永久链接Permalink
resnet_model_cn.html 37.9 KB
Newer Older
T
Deploy to GitHub Pages: 2c54905c3e3154327a8eb57ad96d0770dcc5f435  
Travis CI 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 


<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  
  <title>Model Zoo - ImageNet &mdash; PaddlePaddle  文档</title>
  

  
  

  

  
  
    

  

  
  
    <link rel="stylesheet" href="../../_static/css/theme.css" type="text/css" />
  

  
  
        <link rel="index" title="索引"
              href="../../genindex.html"/>
        <link rel="search" title="搜索" href="../../search.html"/>
    <link rel="top" title="PaddlePaddle  文档" href="../../index.html"/>
        <link rel="up" title="完整教程" href="../index_cn.html"/>
        <link rel="next" title="中文词向量模型的使用" href="../embedding_model/index_cn.html"/>
        <link rel="prev" title="文本生成教程" href="../text_generation/index_cn.html"/> 

  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
  <link rel="stylesheet" href="../../_static/css/override.css" type="text/css" />
  <script>
  var _hmt = _hmt || [];
  (function() {
    var hm = document.createElement("script");
    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
    var s = document.getElementsByTagName("script")[0]; 
    s.parentNode.insertBefore(hm, s);
  })();
  </script>

  

  
  <script src="../../_static/js/modernizr.min.js"></script>

</head>

<body class="wy-body-for-nav" role="document">

  
  <header class="site-header">
    <div class="site-logo">
      <a href="/"><img src="../../_static/images/PP_w.png"></a>
    </div>
    <div class="site-nav-links">
      <div class="site-menu">
        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
        <div class="language-switcher dropdown">
          <a type="button" data-toggle="dropdown">
            <span>English</span>
            <i class="fa fa-angle-up"></i>
            <i class="fa fa-angle-down"></i>
          </a>
          <ul class="dropdown-menu">
            <li><a href="/doc_cn">中文</a></li>
            <li><a href="/doc">English</a></li>
          </ul>
        </div>
        <ul class="site-page-links">
          <li><a>Home</a></li>
          <li><a>Get Started</a></li>
          <li class="active"><a>Documentation</a></li>
          <li><a>About Us</a></li>
        </ul>
      </div>
      <div class="doc-module">
        
        <ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../getstarted/index_cn.html">新手入门</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="../index_cn.html">完整教程</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../howto/index_cn.html">进阶指南</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../api/index_cn.html">API</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../faq/index_cn.html">FAQ</a></li>
</ul>

        
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../../search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>        
      </div>
    </div>
  </header>
  
  <div class="main-content-wrap">

    
    <nav class="doc-menu-vertical" role="navigation">
        
          
          <ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../getstarted/index_cn.html">新手入门</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
T
Deploy to GitHub Pages: 5914c5c7488d48a412c1503ef96de2c1a6b06ed6  
Travis CI 已提交
118 
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
T
Deploy to GitHub Pages: 2c54905c3e3154327a8eb57ad96d0770dcc5f435  
Travis CI 已提交
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 
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../getstarted/basic_usage/index_cn.html">经典的线性回归任务</a></li>
</ul>
</li>
<li class="toctree-l1 current"><a class="reference internal" href="../index_cn.html">完整教程</a><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="../quick_start/index_cn.html">快速入门</a></li>
<li class="toctree-l2"><a class="reference internal" href="../rec/ml_regression_cn.html">个性化推荐</a></li>
<li class="toctree-l2"><a class="reference internal" href="../image_classification/index_cn.html">图像分类</a></li>
<li class="toctree-l2"><a class="reference internal" href="../sentiment_analysis/index_cn.html">情感分析</a></li>
<li class="toctree-l2"><a class="reference internal" href="../semantic_role_labeling/index_cn.html">语义角色标注</a></li>
<li class="toctree-l2"><a class="reference internal" href="../text_generation/index_cn.html">机器翻译</a></li>
<li class="toctree-l2 current"><a class="current reference internal" href="#">ResNet模型</a></li>
<li class="toctree-l2"><a class="reference internal" href="../embedding_model/index_cn.html">词向量模型</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../howto/index_cn.html">进阶指南</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_basis_cn.html#kubernetes">部署Kubernetes集群</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_basis_cn.html#">选择存储方案</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_basis_cn.html#kubectl">配置kubectl</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../howto/deep_model/rnn/rnn_config_cn.html">RNN配置</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
</ul>
</li>
T
Deploy to GitHub Pages: 5f2cbce4190fb83883fed45668b5e4d87a577589  
Travis CI 已提交
164 
<li class="toctree-l1"><a class="reference internal" href="../../api/index_cn.html">API</a><ul>
T
Deploy to GitHub Pages: 362a14d777ec795f1f6b397c1039ebb594050a10  
Travis CI 已提交
165 166 167 168 169 170 171 172 173 174 175 
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/attr.html">Parameter Attribute</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/data.html">数据访问</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/run_logic.html">训练与应用</a></li>
T
Deploy to GitHub Pages: 5f2cbce4190fb83883fed45668b5e4d87a577589  
Travis CI 已提交
176 177 
</ul>
</li>
T
Deploy to GitHub Pages: 2c54905c3e3154327a8eb57ad96d0770dcc5f435  
Travis CI 已提交
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 
<li class="toctree-l1"><a class="reference internal" href="../../faq/index_cn.html">FAQ</a></li>
</ul>

        
    </nav>
    
    <nav class="local-toc"><ul>
<li><a class="reference internal" href="#">Model Zoo - ImageNet</a><ul>
<li><a class="reference internal" href="#resnet">ResNet 介绍</a></li>
<li><a class="reference internal" href="#resnet">ResNet 模型</a><ul>
<li><a class="reference internal" href="#">网络可视化</a></li>
<li><a class="reference internal" href="#">模型下载</a></li>
<li><a class="reference internal" href="#">参数信息</a></li>
<li><a class="reference internal" href="#">参数读取</a></li>
</ul>
</li>
<li><a class="reference internal" href="#">特征提取</a><ul>
<li><a class="reference internal" href="#c">C++接口</a></li>
<li><a class="reference internal" href="#python">Python接口</a></li>
</ul>
</li>
<li><a class="reference internal" href="#">预测</a></li>
</ul>
</li>
</ul>
</nav>
    
    <section class="doc-content-wrap">

      

 







<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
      
        <li><a href="../index_cn.html">完整教程</a> > </li>
      
    <li>Model Zoo - ImageNet</li>
  </ul>
</div>
      
      <div class="wy-nav-content" id="doc-content">
        <div class="rst-content">
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
            
  <div class="section" id="model-zoo-imagenet">
<span id="model-zoo-imagenet"></span><h1>Model Zoo - ImageNet<a class="headerlink" href="#model-zoo-imagenet" title="永久链接至标题"></a></h1>
<p><a class="reference external" href="http://www.image-net.org/">ImageNet</a> 是通用物体分类领域一个众所周知的数据库。本教程提供了一个用于ImageNet上的卷积分类网络模型。</p>
<div class="section" id="resnet">
<span id="resnet"></span><h2>ResNet 介绍<a class="headerlink" href="#resnet" title="永久链接至标题"></a></h2>
<p>论文 <a class="reference external" href="http://arxiv.org/abs/1512.03385">Deep Residual Learning for Image Recognition</a> 中提出的ResNet网络结构在2015年ImageNet大规模视觉识别竞赛(ILSVRC 2015)的分类任务中赢得了第一名。他们提出残差学习的框架来简化网络的训练，所构建网络结构的的深度比之前使用的网络有大幅度的提高。下图展示的是基于残差的连接方式。左图构造网络模块的方式被用于34层的网络中，而右图的瓶颈连接模块用于50层，101层和152层的网络结构中。</p>
<p><center><img alt="resnet_block" src="../../_images/resnet_block.jpg" /></center>
<center>图 1. ResNet 网络模块</center></p>
<p>本教程中我们给出了三个ResNet模型，这些模型都是由原作者提供的模型<a class="reference external" href="https://github.com/KaimingHe/deep-residual-networks">https://github.com/KaimingHe/deep-residual-networks</a>转换过来的。我们使用PaddlePaddle在ILSVRC的验证集共50,000幅图像上测试了模型的分类错误率，其中输入图像的颜色通道顺序为<strong>BGR</strong>，保持宽高比缩放到短边为256，只截取中心方形的图像区域。分类错误率和模型大小由下表给出。
<center></p>
<table border="2" cellspacing="0" cellpadding="6" rules="all" frame="border">
<colgroup>
<col  class="left" />
<col  class="left" />
<col  class="left" />
</colgroup>
<thead>
<tr>
<th scope="col" class="left">ResNet</th>
<th scope="col" class="left">Top-1</th>
<th scope="col" class="left">Model Size</th>
</tr>
</thead><tbody>
<tr>
<td class="left">ResNet-50</td>
<td class="left">24.9%</td>
<td class="left">99M</td>
</tr>
<tr>
<td class="left">ResNet-101</td>
<td class="left">23.7%</td>
<td class="left">173M</td>
</tr>
<tr>
<td class="left">ResNet-152</td>
<td class="left">23.2%</td>
<td class="left">234M</td>
</tr>
</tbody></table></center>
<br></div>
<div class="section" id="resnet">
<span id="id1"></span><h2>ResNet 模型<a class="headerlink" href="#resnet" title="永久链接至标题"></a></h2>
<p>50层，101层和152层的网络配置文件可参照<code class="docutils literal"><span class="pre">demo/model_zoo/resnet/resnet.py</span></code>。你也可以通过在命令行参数中增加一个参数如<code class="docutils literal"><span class="pre">--config_args=layer_num=50</span></code>来指定网络层的数目。</p>
<div class="section" id="">
<span id="id2"></span><h3>网络可视化<a class="headerlink" href="#" title="永久链接至标题"></a></h3>
<p>你可以通过执行下面的命令来得到ResNet网络的结构可视化图。该脚本会生成一个dot文件，然后可以转换为图片。需要安装graphviz来转换dot文件为图片。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">cd</span> <span class="n">demo</span><span class="o">/</span><span class="n">model_zoo</span><span class="o">/</span><span class="n">resnet</span>
<span class="o">./</span><span class="n">net_diagram</span><span class="o">.</span><span class="n">sh</span>
</pre></div>
</div>
</div>
<div class="section" id="">
<span id="id3"></span><h3>模型下载<a class="headerlink" href="#" title="永久链接至标题"></a></h3>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">cd</span> <span class="n">demo</span><span class="o">/</span><span class="n">model_zoo</span><span class="o">/</span><span class="n">resnet</span>
<span class="o">./</span><span class="n">get_model</span><span class="o">.</span><span class="n">sh</span>
</pre></div>
</div>
<p>你可以执行上述命令来下载所有的模型和均值文件，如果下载成功，这些文件将会被保存在<code class="docutils literal"><span class="pre">demo/model_zoo/resnet/model</span></code>路径下。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">mean_meta_224</span>  <span class="n">resnet_101</span>  <span class="n">resnet_152</span>  <span class="n">resnet_50</span>
</pre></div>
</div>
<ul class="simple">
<li>resnet_50: 50层网络模型。</li>
<li>resnet_101: 101层网络模型。</li>
<li>resnet_152: 152层网络模型。</li>
<li>mean_meta_224: 均值图像文件，图像大小为3 x 224 x 224，颜色通道顺序为<strong>BGR</strong>。你也可以使用这三个值: 103.939, 116.779, 123.68。</li>
</ul>
</div>
<div class="section" id="">
<span id="id4"></span><h3>参数信息<a class="headerlink" href="#" title="永久链接至标题"></a></h3>
<ul>
<li><p class="first"><strong>卷积层权重</strong></p>
<p>由于每个卷积层后面连接的是batch normalization层，因此该层中没有偏置(bias)参数，并且只有一个权重。
形状: <code class="docutils literal"><span class="pre">(Co,</span> <span class="pre">ky,</span> <span class="pre">kx,</span> <span class="pre">Ci)</span></code></p>
<ul class="simple">
<li>Co: 输出特征图的通道数目</li>
<li>ky: 滤波器核在垂直方向上的尺寸</li>
<li>kx: 滤波器核在水平方向上的尺寸</li>
<li>Ci: 输入特征图的通道数目</li>
</ul>
<p>二维矩阵: (Co * ky * kx, Ci), 行优先次序存储。</p>
</li>
<li><p class="first"><strong>全连接层权重</strong></p>
<p>二维矩阵: (输入层尺寸, 本层尺寸), 行优先次序存储。</p>
</li>
<li><p class="first"><strong><a class="reference external" href="http://arxiv.org/abs/1502.03167">Batch Normalization</a> 层权重</strong></p>
</li>
</ul>
<p>本层有四个参数，实际上只有.w0和.wbias是需要学习的参数，另外两个分别是滑动均值和方差。在测试阶段它们将会被加载到模型中。下表展示了batch normalization层的参数。
<center></p>
<table border="2" cellspacing="0" cellpadding="6" rules="all" frame="border">
<colgroup>
<col  class="left" />
<col  class="left" />
<col  class="left" />
</colgroup>
<thead>
<tr>
<th scope="col" class="left">参数名</th>
<th scope="col" class="left">尺寸</th>
<th scope="col" class="left">含义</th>
</tr>
</thead><tbody>
<tr>
<td class="left">_res2_1_branch1_bn.w0</td>
<td class="left">256</td>
<td class="left">gamma, 缩放参数</td>
</tr>
<tr>
<td class="left">_res2_1_branch1_bn.w1</td>
<td class="left">256</td>
<td class="left">特征图均值</td>
</tr>
<tr>
<td class="left">_res2_1_branch1_bn.w2</td>
<td class="left">256</td>
<td class="left">特征图方差</td>
</tr>
<tr>
<td class="left">_res2_1_branch1_bn.wbias</td>
<td class="left">256</td>
<td class="left">beta, 偏置参数</td>
</tr>
</tbody></table></center>
<br></div>
<div class="section" id="">
<span id="id5"></span><h3>参数读取<a class="headerlink" href="#" title="永久链接至标题"></a></h3>
<p>使用者可以使用下面的Python脚本来读取参数值:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">sys</span>
<span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>

<span class="k">def</span> <span class="nf">load</span><span class="p">(</span><span class="n">file_name</span><span class="p">):</span>
    <span class="k">with</span> <span class="nb">open</span><span class="p">(</span><span class="n">file_name</span><span class="p">,</span> <span class="s1">&#39;rb&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">f</span><span class="p">:</span>
        <span class="n">f</span><span class="o">.</span><span class="n">read</span><span class="p">(</span><span class="mi">16</span><span class="p">)</span> <span class="c1"># skip header for float type.</span>
        <span class="k">return</span> <span class="n">np</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="n">f</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">float32</span><span class="p">)</span>

<span class="k">if</span> <span class="vm">__name__</span><span class="o">==</span><span class="s1">&#39;__main__&#39;</span><span class="p">:</span>
    <span class="n">weight</span> <span class="o">=</span> <span class="n">load</span><span class="p">(</span><span class="n">sys</span><span class="o">.</span><span class="n">argv</span><span class="p">[</span><span class="mi">1</span><span class="p">])</span>
</pre></div>
</div>
<p>或者直接使用下面的shell命令:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">od</span> <span class="o">-</span><span class="n">j</span> <span class="mi">16</span> <span class="o">-</span><span class="n">f</span> <span class="n">_res2_1_branch1_bn</span><span class="o">.</span><span class="n">w0</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="">
<span id="id6"></span><h2>特征提取<a class="headerlink" href="#" title="永久链接至标题"></a></h2>
<p>我们提供了C++和Python接口来提取特征。下面的例子使用了<code class="docutils literal"><span class="pre">demo/model_zoo/resnet/example</span></code>中的数据，详细地展示了整个特征提取的过程。</p>
<div class="section" id="c">
<span id="c"></span><h3>C++接口<a class="headerlink" href="#c" title="永久链接至标题"></a></h3>
<p>首先，在配置文件中的<code class="docutils literal"><span class="pre">define_py_data_sources2</span></code>里指定图像数据列表，具体请参照示例<code class="docutils literal"><span class="pre">demo/model_zoo/resnet/resnet.py</span></code>。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span>    <span class="n">train_list</span> <span class="o">=</span> <span class="s1">&#39;train.list&#39;</span> <span class="k">if</span> <span class="ow">not</span> <span class="n">is_test</span> <span class="k">else</span> <span class="kc">None</span>
    <span class="c1"># mean.meta is mean file of ImageNet dataset.</span>
    <span class="c1"># mean.meta size : 3 x 224 x 224.</span>
    <span class="c1"># If you use three mean value, set like:</span>
    <span class="c1"># &quot;mean_value:103.939,116.779,123.68;&quot;</span>
    <span class="n">args</span><span class="o">=</span><span class="p">{</span>
        <span class="s1">&#39;mean_meta&#39;</span><span class="p">:</span> <span class="s2">&quot;model/mean_meta_224/mean.meta&quot;</span><span class="p">,</span>
        <span class="s1">&#39;image_size&#39;</span><span class="p">:</span> <span class="mi">224</span><span class="p">,</span> <span class="s1">&#39;crop_size&#39;</span><span class="p">:</span> <span class="mi">224</span><span class="p">,</span>
        <span class="s1">&#39;color&#39;</span><span class="p">:</span> <span class="kc">True</span><span class="p">,</span><span class="s1">&#39;swap_channel:&#39;</span><span class="p">:</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">]}</span>
    <span class="n">define_py_data_sources2</span><span class="p">(</span><span class="n">train_list</span><span class="p">,</span>
                           <span class="s1">&#39;example/test.list&#39;</span><span class="p">,</span>
                           <span class="n">module</span><span class="o">=</span><span class="s2">&quot;example.image_list_provider&quot;</span><span class="p">,</span>
                           <span class="n">obj</span><span class="o">=</span><span class="s2">&quot;processData&quot;</span><span class="p">,</span>
                           <span class="n">args</span><span class="o">=</span><span class="n">args</span><span class="p">)</span>
</pre></div>
</div>
<p>第二步，在<code class="docutils literal"><span class="pre">resnet.py</span></code>文件中指定要提取特征的网络层的名字。例如，</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">Outputs</span><span class="p">(</span><span class="s2">&quot;res5_3_branch2c_conv&quot;</span><span class="p">,</span> <span class="s2">&quot;res5_3_branch2c_bn&quot;</span><span class="p">)</span>
</pre></div>
</div>
<p>第三步，在<code class="docutils literal"><span class="pre">extract_fea_c++.sh</span></code>文件中指定模型路径和输出的目录，然后执行下面的命令。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">cd</span> <span class="n">demo</span><span class="o">/</span><span class="n">model_zoo</span><span class="o">/</span><span class="n">resnet</span>
<span class="o">./</span><span class="n">extract_fea_c</span><span class="o">++.</span><span class="n">sh</span>
</pre></div>
</div>
<p>如果执行成功，特征将会存到<code class="docutils literal"><span class="pre">fea_output/rank-00000</span></code>文件中，如下所示。同时你可以使用<code class="docutils literal"><span class="pre">load_feature.py</span></code>文件中的<code class="docutils literal"><span class="pre">load_feature_c</span></code>接口来加载该文件。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">-</span><span class="mf">0.115318</span> <span class="o">-</span><span class="mf">0.108358</span> <span class="o">...</span> <span class="o">-</span><span class="mf">0.087884</span><span class="p">;</span><span class="o">-</span><span class="mf">1.27664</span> <span class="o">...</span> <span class="o">-</span><span class="mf">1.11516</span> <span class="o">-</span><span class="mf">2.59123</span><span class="p">;</span>
<span class="o">-</span><span class="mf">0.126383</span> <span class="o">-</span><span class="mf">0.116248</span> <span class="o">...</span> <span class="o">-</span><span class="mf">0.00534909</span><span class="p">;</span><span class="o">-</span><span class="mf">1.42593</span> <span class="o">...</span> <span class="o">-</span><span class="mf">1.04501</span> <span class="o">-</span><span class="mf">1.40769</span><span class="p">;</span>
</pre></div>
</div>
<ul class="simple">
<li>每行存储的是一个样本的特征。其中，第一行存的是图像<code class="docutils literal"><span class="pre">example/dog.jpg</span></code>的特征，第二行存的是图像<code class="docutils literal"><span class="pre">example/cat.jpg</span></code>的特征。</li>
<li>不同层的特征由分号<code class="docutils literal"><span class="pre">;</span></code>隔开，并且它们的顺序与<code class="docutils literal"><span class="pre">Outputs()</span></code>中指定的层顺序一致。这里，左边是<code class="docutils literal"><span class="pre">res5_3_branch2c_conv</span></code>层的特征，右边是<code class="docutils literal"><span class="pre">res5_3_branch2c_bn</span></code>层特征。</li>
</ul>
</div>
<div class="section" id="python">
<span id="python"></span><h3>Python接口<a class="headerlink" href="#python" title="永久链接至标题"></a></h3>
<p>示例<code class="docutils literal"><span class="pre">demo/model_zoo/resnet/classify.py</span></code>中展示了如何使用Python来提取特征。下面的例子同样使用了<code class="docutils literal"><span class="pre">./example/test.list</span></code>中的数据。执行的命令如下：</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">cd</span> <span class="n">demo</span><span class="o">/</span><span class="n">model_zoo</span><span class="o">/</span><span class="n">resnet</span>
<span class="o">./</span><span class="n">extract_fea_py</span><span class="o">.</span><span class="n">sh</span>
</pre></div>
</div>
<p>extract_fea_py.sh:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">python</span> <span class="n">classify</span><span class="o">.</span><span class="n">py</span> \
     <span class="o">--</span><span class="n">job</span><span class="o">=</span><span class="n">extract</span> \
     <span class="o">--</span><span class="n">conf</span><span class="o">=</span><span class="n">resnet</span><span class="o">.</span><span class="n">py</span>\
     <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span> \
     <span class="o">--</span><span class="n">mean</span><span class="o">=</span><span class="n">model</span><span class="o">/</span><span class="n">mean_meta_224</span><span class="o">/</span><span class="n">mean</span><span class="o">.</span><span class="n">meta</span> \
     <span class="o">--</span><span class="n">model</span><span class="o">=</span><span class="n">model</span><span class="o">/</span><span class="n">resnet_50</span> \
     <span class="o">--</span><span class="n">data</span><span class="o">=./</span><span class="n">example</span><span class="o">/</span><span class="n">test</span><span class="o">.</span><span class="n">list</span> \
     <span class="o">--</span><span class="n">output_layer</span><span class="o">=</span><span class="s2">&quot;res5_3_branch2c_conv,res5_3_branch2c_bn&quot;</span> \
     <span class="o">--</span><span class="n">output_dir</span><span class="o">=</span><span class="n">features</span>
</pre></div>
</div>
<ul class="simple">
<li>--job=extract:              指定工作模式来提取特征。</li>
<li>--conf=resnet.py:           网络配置文件。</li>
<li>--use_gpu=1:                指定是否使用GPU。</li>
<li>--model=model/resnet_50:    模型路径。</li>
<li>--data=./example/test.list: 数据列表。</li>
<li>--output_layer=&#8221;xxx,xxx&#8221;:   指定提取特征的层。</li>
<li>--output_dir=features:      输出目录。</li>
</ul>
<p>如果运行成功，你将会看到特征存储在<code class="docutils literal"><span class="pre">features/batch_0</span></code>文件中，该文件是由cPickle产生的。你可以使用<code class="docutils literal"><span class="pre">load_feature.py</span></code>中的<code class="docutils literal"><span class="pre">load_feature_py</span></code>接口来打开该文件，它将返回如下的字典：</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="p">{</span>
<span class="s1">&#39;cat.jpg&#39;</span><span class="p">:</span> <span class="p">{</span><span class="s1">&#39;res5_3_branch2c_conv&#39;</span><span class="p">:</span> <span class="n">array</span><span class="p">([[</span><span class="o">-</span><span class="mf">0.12638293</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.116248</span>  <span class="p">,</span> <span class="o">-</span><span class="mf">0.11883899</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.00895038</span><span class="p">,</span> <span class="mf">0.01994277</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.00534909</span><span class="p">]],</span> <span class="n">dtype</span><span class="o">=</span><span class="n">float32</span><span class="p">),</span> <span class="s1">&#39;res5_3_branch2c_bn&#39;</span><span class="p">:</span> <span class="n">array</span><span class="p">([[</span><span class="o">-</span><span class="mf">1.42593431</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.28918779</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.32414699</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.45933616</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.04501402</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.40769434</span><span class="p">]],</span> <span class="n">dtype</span><span class="o">=</span><span class="n">float32</span><span class="p">)},</span>
<span class="s1">&#39;dog.jpg&#39;</span><span class="p">:</span> <span class="p">{</span><span class="s1">&#39;res5_3_branch2c_conv&#39;</span><span class="p">:</span> <span class="n">array</span><span class="p">([[</span><span class="o">-</span><span class="mf">0.11531784</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.10835785</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.08809858</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span><span class="mf">0.0055237</span><span class="p">,</span> <span class="mf">0.01505112</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.08788397</span><span class="p">]],</span> <span class="n">dtype</span><span class="o">=</span><span class="n">float32</span><span class="p">),</span> <span class="s1">&#39;res5_3_branch2c_bn&#39;</span><span class="p">:</span> <span class="n">array</span><span class="p">([[</span><span class="o">-</span><span class="mf">1.27663755</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.18272924</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.90937918</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.25178063</span><span class="p">,</span> <span class="o">-</span><span class="mf">1.11515927</span><span class="p">,</span> <span class="o">-</span><span class="mf">2.59122872</span><span class="p">]],</span> <span class="n">dtype</span><span class="o">=</span><span class="n">float32</span><span class="p">)}</span>
<span class="p">}</span>
</pre></div>
</div>
<p>仔细观察，这些特征值与上述使用C++接口提取的结果是一致的。</p>
</div>
</div>
<div class="section" id="">
<span id="id7"></span><h2>预测<a class="headerlink" href="#" title="永久链接至标题"></a></h2>
<p><code class="docutils literal"><span class="pre">classify.py</span></code>文件也可以用于对样本进行预测。我们提供了一个示例脚本<code class="docutils literal"><span class="pre">predict.sh</span></code>，它使用50层的ResNet模型来对<code class="docutils literal"><span class="pre">example/test.list</span></code>中的数据进行预测。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">cd</span> <span class="n">demo</span><span class="o">/</span><span class="n">model_zoo</span><span class="o">/</span><span class="n">resnet</span>
<span class="o">./</span><span class="n">predict</span><span class="o">.</span><span class="n">sh</span>
</pre></div>
</div>
<p>predict.sh调用了<code class="docutils literal"><span class="pre">classify.py</span></code>:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">python</span> <span class="n">classify</span><span class="o">.</span><span class="n">py</span> \
     <span class="o">--</span><span class="n">job</span><span class="o">=</span><span class="n">predict</span> \
     <span class="o">--</span><span class="n">conf</span><span class="o">=</span><span class="n">resnet</span><span class="o">.</span><span class="n">py</span>\
     <span class="o">--</span><span class="n">multi_crop</span> \
     <span class="o">--</span><span class="n">model</span><span class="o">=</span><span class="n">model</span><span class="o">/</span><span class="n">resnet_50</span> \
     <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span> \
     <span class="o">--</span><span class="n">data</span><span class="o">=./</span><span class="n">example</span><span class="o">/</span><span class="n">test</span><span class="o">.</span><span class="n">list</span>
</pre></div>
</div>
<ul class="simple">
<li>--job=extract:              指定工作模型进行预测。</li>
<li>--conf=resnet.py:           网络配置文件。network configure.</li>
<li>--multi_crop:               使用10个裁剪图像块，预测概率取平均。</li>
<li>--use_gpu=1:                指定是否使用GPU。</li>
<li>--model=model/resnet_50:    模型路径。</li>
<li>--data=./example/test.list: 数据列表。</li>
</ul>
<p>如果运行成功，你将会看到如下结果，其中156和285是这些图像的分类标签。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">Label</span> <span class="n">of</span> <span class="n">example</span><span class="o">/</span><span class="n">dog</span><span class="o">.</span><span class="n">jpg</span> <span class="ow">is</span><span class="p">:</span> <span class="mi">156</span>
<span class="n">Label</span> <span class="n">of</span> <span class="n">example</span><span class="o">/</span><span class="n">cat</span><span class="o">.</span><span class="n">jpg</span> <span class="ow">is</span><span class="p">:</span> <span class="mi">282</span>
</pre></div>
</div>
</div>
</div>


           </div>
          </div>
          <footer>
  
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
      
        <a href="../embedding_model/index_cn.html" class="btn btn-neutral float-right" title="中文词向量模型的使用" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
      
      
        <a href="../text_generation/index_cn.html" class="btn btn-neutral" title="文本生成教程" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
      
    </div>
  

  <hr/>

  <div role="contentinfo">
    <p>
        &copy; Copyright 2016, PaddlePaddle developers.

    </p>
  </div>
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 

</footer>

        </div>
      </div>

    </section>

  </div>
  


  

    <script type="text/javascript">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'../../',
            VERSION:'',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true
        };
    </script>
      <script type="text/javascript" src="../../_static/jquery.js"></script>
      <script type="text/javascript" src="../../_static/underscore.js"></script>
      <script type="text/javascript" src="../../_static/doctools.js"></script>
      <script type="text/javascript" src="../../_static/translations.js"></script>
      <script type="text/javascript" src="https://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
       
  

  
  
    <script type="text/javascript" src="../../_static/js/theme.js"></script>
  
  
  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
  <script src="../../_static/js/paddle_doc_init.js"></script> 

</body>
</html>