3.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import tensorflow as tf\n",
      "import numpy as np\n",
      "\n",
      "# Create 100 phony x, y data points in NumPy, y = x * 0.1 + 0.3\n",
      "x_data = np.random.rand(100).astype(np.float32)\n",
      "y_data = x_data * 0.1 + 0.3\n",
      "\n",
      "# Try to find values for W and b that compute y_data = W * x_data + b\n",
      "# (We know that W should be 0.1 and b 0.3, but Tensorflow will\n",
      "# figure that out for us.)\n",
      "W = tf.Variable(tf.random_uniform([1], -1.0, 1.0))\n",
      "b = tf.Variable(tf.zeros([1]))\n",
      "y = W * x_data + b\n",
      "\n",
      "# Minimize the mean squared errors.\n",
      "loss = tf.reduce_mean(tf.square(y - y_data))\n",
      "optimizer = tf.train.GradientDescentOptimizer(0.5)\n",
      "train = optimizer.minimize(loss)\n",
      "\n",
      "# Before starting, initialize the variables.  We will 'run' this first.\n",
      "init = tf.initialize_all_variables()\n",
      "\n",
      "# Launch the graph.\n",
      "sess = tf.Session()\n",
      "sess.run(init)\n",
      "\n",
      "# Fit the line.\n",
      "for step in xrange(201):\n",
      "    sess.run(train)\n",
      "    if step % 20 == 0:\n",
      "        print(step, sess.run(W), sess.run(b))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(0, array([ 0.52948135], dtype=float32), array([ 0.10670546], dtype=float32))\n",
        "(20, array([ 0.20451508], dtype=float32), array([ 0.24768613], dtype=float32))\n",
        "(40, array([ 0.12595835], dtype=float32), array([ 0.28700686], dtype=float32))\n",
        "(60, array([ 0.10644727], dtype=float32), array([ 0.2967729], dtype=float32))\n",
        "(80, array([ 0.10160133], dtype=float32), array([ 0.29919848], dtype=float32))\n",
        "(100, array([ 0.10039775], dtype=float32), array([ 0.29980093], dtype=float32))\n",
        "(120, array([ 0.10009881], dtype=float32), array([ 0.29995054], dtype=float32))\n",
        "(140, array([ 0.10002455], dtype=float32), array([ 0.29998773], dtype=float32))\n",
        "(160, array([ 0.1000061], dtype=float32), array([ 0.29999694], dtype=float32))"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "(180, array([ 0.10000155], dtype=float32), array([ 0.29999924], dtype=float32))\n",
        "(200, array([ 0.1000004], dtype=float32), array([ 0.2999998], dtype=float32))\n"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import tensorflow as tf\n",
      "sess = tf.InteractiveSession()\n",
      "x = tf.Variable([1.0, 2.0])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import csv\n",
      "\n",
      "csvfile = file('csv_test.csv', 'wb')\n",
      "writer = csv.writer(csvfile)\n",
      "writer.writerow(['\u59d3\u540d', '\u5e74\u9f84', '\u7535\u8bdd'])\n",
      "\n",
      "data = [range(3) for i in range(3)]\n",
      "for item in data:\n",
      "    writer.writerows(item)\n",
      "\n",
      "csvfile.close()\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "ename": "Error",
       "evalue": "sequence expected",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[0;31mError\u001b[0m                                     Traceback (most recent call last)",
        "\u001b[0;32m<ipython-input-9-81d1a629a305>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mitem\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 9\u001b[0;31m     \u001b[0mwriter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwriterows\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     10\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m \u001b[0mcsvfile\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
        "\u001b[0;31mError\u001b[0m: sequence expected"
       ]
      }
     ],
     "prompt_number": 9
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = [range(3) for i in range(3)]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "[[0, 1, 2], [0, 1, 2], [0, 1, 2]]"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "import cv2\n",
      "import sys\n",
      "import math\n",
      "import pickle\n",
      "import tensorflow as tf\n",
      "def GetVideoProperties(filePath):\n",
      "    inputVideo = cv2.VideoCapture(filePath);\n",
      "    fileName = filePath\n",
      "    if inputVideo.isOpened() == None:\n",
      "        return False;\n",
      "    frameCount = 0;\n",
      "    while inputVideo.isOpened():\n",
      "        ret, frame = inputVideo.read()\n",
      "        if ret == False:\n",
      "            break;\n",
      "        if frame is None:\n",
      "            return False;\n",
      "        if frame.shape[2] != 3:\n",
      "            return False;\n",
      "        frameSize = (int(inputVideo.get(cv2.CAP_PROP_FRAME_WIDTH)), int(inputVideo.get(cv2.CAP_PROP_FRAME_HEIGHT)))\n",
      "        frameCount += 1;\n",
      "    fps = inputVideo.get(cv2.CAP_PROP_FPS)\n",
      "    return [fps, frameSize, frameCount]\n",
      "\n",
      "def NumToNNValuesSimple(num, maxValue):\n",
      "    res = np.zeros((maxValue, 1))\n",
      "    res[num, 0]= 1.0\n",
      "    res = res.flatten()\n",
      "    return res\n",
      "\n",
      "def ImageToNNValues(img):\n",
      "    #print img\n",
      "    result = img.flatten()\n",
      "    result = result / 256.0\n",
      "    return result\n",
      "\n",
      "def NNValuesToImage(values, frameSize):\n",
      "    values *= 256.0\n",
      "    result = values.reshape(frameSize[0], frameSize[1], 3)\n",
      "    return result\n",
      "\n",
      "def ReconstructFrame(nnoutput, frameSize, frameNum, frameCount, NumToNNValuesSimple):\n",
      "    frame = NNValuesToImage(nnoutput, frameSize)\n",
      "    return frame\n",
      "\n",
      "def ReconstructMovie(nnoutput, frameSize, frameCount, fps, filePath, NumToNNValuesSimple):\n",
      "    \n",
      "    print \"Loading neural network\"\n",
      "    if nnoutput is None:\n",
      "        return False\n",
      "    fourcc = cv2.VideoWriter_fourcc(*'XVID');\n",
      "    outputVideo = cv2.VideoWriter(filePath, fourcc, fps, frameSize);\n",
      "    if outputVideo.isOpened() == None:\n",
      "        print \"Could not open the output video for write: \", filePath\n",
      "        return False;\n",
      "    for frameNum in range(frameCount):\n",
      "        if  frameNum == 0 or frameNum == frameCount - 1 or frameNum % 100 == 0:\n",
      "            print \"Reconstructing frame \", frameNum + 1, \" of \", frameCount\n",
      "        frame = ReconstructFrame(nnoutput[frameNum], frameSize, frameNum, frameCount, NumToNNValuesSimple)\n",
      "        #norm = np.linalg.norm(frame)\n",
      "        #frame = frame / norm\n",
      "        #print frame\n",
      "        #frame = frame * 256.0\n",
      "        frame = frame.astype('u1')\n",
      "        \n",
      "        #print len(frame), len(frame[0]), len(frame[0][0])\n",
      "        cv2.imshow('frame',frame)\n",
      "        if cv2.waitKey(1) & 0xFF == ord('q'):\n",
      "            break\n",
      "        outputVideo.write(frame);\n",
      "    print \"Saved \", filePath\n",
      "    return True\n",
      "\n",
      "if __name__ == '__main__':\n",
      "\n",
      "    inFilePath = \"/home/lenovo3/3/box1.avi\"\n",
      "\n",
      "    nnFilePath = inFilePath + \".nncv\";\n",
      "    outFilePath = inFilePath + \".n_n.avi\";\n",
      "    maxIters = 1000;\n",
      "    epsilon = 0.00000000001;\n",
      "\n",
      "    [fps, frameSize, frameCount] = GetVideoProperties(inFilePath)\n",
      "    \n",
      "    print inFilePath, \" - fps: \", fps, \" - frameSize: \", frameSize, \" - frameCount: \", frameCount\n",
      "\n",
      "    layerSizes = [];\n",
      "    \n",
      "    inputLayerSize = NumToNNValuesSimple(0, frameCount).shape[0];\n",
      "    outputLayerSize = frameSize[0] * frameSize[1] * 3;\n",
      "    hiddenLayerSize = int(math.sqrt(frameCount)) + 1;\n",
      "    layerSizes.append(inputLayerSize );\n",
      "    layerSizes.append(hiddenLayerSize );\n",
      "    layerSizes.append(hiddenLayerSize );\n",
      "    layerSizes.append(outputLayerSize );\n",
      "    layerSizes = np.array(layerSizes)\n",
      "    \n",
      "    nnPtr = cv2.ml.ANN_MLP_create();\n",
      "    nnPtr.setBackpropMomentumScale(0.0)\n",
      "    nnPtr.setBackpropWeightScale(0.001)\n",
      "    nnPtr.setLayerSizes( layerSizes );\n",
      "    nnPtr.setActivationFunction(cv2.ml.ANN_MLP_GAUSSIAN, 2, 1);\n",
      "    nnPtr.setTrainMethod( cv2.ml.ANN_MLP_RPROP, 0.1, sys.float_info.epsilon)\n",
      "    nnPtr.setTermCriteria((cv2.TERM_CRITERIA_MAX_ITER + cv2.TERM_CRITERIA_EPS, maxIters, epsilon));\n",
      "\n",
      "    samples = np.empty((frameCount, inputLayerSize))\n",
      "    responses = np.empty((frameCount, outputLayerSize))\n",
      "\n",
      "    inputVideo = cv2.VideoCapture(inFilePath) ;\n",
      "    fileName = inFilePath;\n",
      "\n",
      "    for frameNum in range(frameCount):\n",
      "        if frameNum == 0 or frameNum == frameCount - 1 or frameNum % 100 == 0:\n",
      "            print \"Loading frame \", frameNum + 1, \" of \", frameCount\n",
      "        ret, frame = inputVideo.read()\n",
      "        if ret == False:\n",
      "            break;\n",
      "        if frame is None:\n",
      "            print 1;\n",
      "        imageNNValues = ImageToNNValues(frame);\n",
      "        frameNumNNValues = NumToNNValuesSimple(frameNum, frameCount);\n",
      "        #print len(samples), len(samples[0]), len(responses), len(responses[0])\n",
      "        np.copyto(samples[frameNum], frameNumNNValues);\n",
      "        np.copyto(responses[frameNum], imageNNValues);\n",
      "\n",
      "    def weight_variable(shape):\n",
      "      initial = tf.truncated_normal(shape, stddev=0.1)\n",
      "      return tf.Variable(initial)\n",
      "    \n",
      "    def bias_variable(shape):\n",
      "      initial = tf.constant(0.1, shape=shape)\n",
      "      return tf.Variable(initial)\n",
      "    \n",
      "    def conv2d(x, W):\n",
      "      return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')\n",
      "    \n",
      "    def max_pool_2x2(x):\n",
      "      return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],\n",
      "                            strides=[1, 2, 2, 1], padding='SAME')\n",
      "    \n",
      "    frameCount = 96\n",
      "    x = tf.placeholder(\"float\", shape=[None, 96])\n",
      "    y_ = tf.placeholder(\"float\", shape=[None, frameSize[0] * frameSize[1] * 3])\n",
      "    \n",
      "    W_fc1 = weight_variable([frameCount, 1024])\n",
      "    b_fc1 = bias_variable([1024])\n",
      "    \n",
      "    #print tf.rank(x_vector), tf.rank(W_fc1), tf.rank(b_fc1)\n",
      "    h_fc1 = tf.nn.softmax(tf.matmul(x, W_fc1) + b_fc1)\n",
      "    \n",
      "    W_fc1 = weight_variable([1024,frameCount * frameCount])\n",
      "    b_fc1 = bias_variable([frameCount*frameCount])\n",
      "    \n",
      "    h_flat2 = tf.reshape(h_fc1, [-1, 1024])\n",
      "    h_fc2 = tf.nn.relu(tf.matmul(h_flat2, W_fc1) + b_fc1)\n",
      "    \n",
      "    h_fc2 = tf.reshape(h_fc2, [1, frameCount, frameCount, 32])\n",
      "    print tf.shape(h_fc2)\n",
      "    \n",
      "    W_conv2 = weight_variable([10, 10, 32, 64])\n",
      "    b_conv2 = bias_variable([64])\n",
      "    \n",
      "    h_conv2 = tf.nn.relu(conv2d(h_fc2, W_conv2) + b_conv2)\n",
      "    h_pool2 = max_pool_2x2(h_conv2)\n",
      "    \n",
      "    W_conv1 = weight_variable([5, 5, 32, 2700])\n",
      "    b_conv1 = bias_variable([2700])\n",
      "    \n",
      "    \n",
      "    x_image = tf.reshape(h_pool2, [-1,frameCount,frameCount,1])\n",
      "    h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)\n",
      "    \n",
      "    y_pool = max_pool_2x2(h_conv1)\n",
      "    \n",
      "    cross_entropy = -tf.reduce_sum(y_*tf.log(y_pool))\n",
      "    train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)\n",
      "    correct_prediction = tf.equal(tf.argmax(y_pool,1), tf.argmax(y_,1))\n",
      "    accuracy = tf.reduce_mean(tf.cast(correct_prediction, \"float\"))\n",
      "    sess.run(tf.initialize_all_variables())\n",
      "    for i in range(2000):\n",
      "      batch_xs, batch_ys = samples, responses\n",
      "      if i%100 == 0:\n",
      "        train_accuracy = accuracy.eval(feed_dict={x:batch_xs, y_: batch_ys})\n",
      "        print \"step %d, training accuracy %g\"%(i, train_accuracy)\n",
      "      train_step.run(feed_dict={x: batch_xs, y_: batch_ys})\n",
      "    \n",
      "    print \"test accuracy %g\"%accuracy.eval(feed_dict={x: samples, y_: responses})"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "ename": "UnboundLocalError",
       "evalue": "local variable 'frameSize' referenced before assignment",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[0;31mUnboundLocalError\u001b[0m                         Traceback (most recent call last)",
        "\u001b[0;32m<ipython-input-2-4dfba7277404>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     82\u001b[0m     \u001b[0mepsilon\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0.00000000001\u001b[0m\u001b[0;34m;\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     83\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 84\u001b[0;31m     \u001b[0;34m[\u001b[0m\u001b[0mfps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeSize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeCount\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mGetVideoProperties\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minFilePath\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     85\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     86\u001b[0m     \u001b[0;32mprint\u001b[0m \u001b[0minFilePath\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\" - fps: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\" - frameSize: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeSize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\" - frameCount: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeCount\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
        "\u001b[0;32m<ipython-input-2-4dfba7277404>\u001b[0m in \u001b[0;36mGetVideoProperties\u001b[0;34m(filePath)\u001b[0m\n\u001b[1;32m     22\u001b[0m         \u001b[0mframeCount\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m;\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m     \u001b[0mfps\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0minputVideo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcv2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mCAP_PROP_FPS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 24\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mfps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeSize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mframeCount\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     25\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     26\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mNumToNNValuesSimple\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnum\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmaxValue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
        "\u001b[0;31mUnboundLocalError\u001b[0m: local variable 'frameSize' referenced before assignment"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}