960 lines (959 with data), 196.7 kB

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from mylib.utils.misc import plot_voxel_enhance\n",
    "from mylib.dataloader import dataset\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf_seg_dataset = dataset.ClfSegDataset(crop_size=32, subset=[0,3,4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(clf_seg_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x, (y, seg) = clf_seg_dataset[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(32, 32, 32, 1)\n",
      "[True, False, False]\n",
      "(32, 32, 32, 1)\n"
     ]
    }
   ],
   "source": [
    "print(x.shape)\n",
    "print(y)\n",
    "print(seg.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAksAAAJCCAYAAADQsoPKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0b00dcb2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_voxel_enhance(x.squeeze(), seg.squeeze(), alpha=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The interactive 3D nodule mesh plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# from IPython.display import HTML\n",
    "# from mylib.utils.plot3d import plotly_3d_scan_to_html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# plotly_3d_scan_to_html(seg.squeeze(),\"tmp.html\",step_size=3, zyx_range=((0,32),)*3)\n",
    "# HTML(\"tmp.html\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create the models * 3D DenseSharp* and * 3D DenseNet*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jiancheng/anaconda3/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6\n",
      "  return f(*args, **kwds)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TensorFlow version 1.4.0\n",
      "Keras version 2.1.5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import keras\n",
    "print(\"TensorFlow version\",tf.__version__)\n",
    "print(\"Keras version\",keras.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from mylib.models.misc import set_gpu_usage\n",
    "set_gpu_usage()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from mylib.models import densenet, densesharp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model hyper-parameters: {'activation': <function <lambda> at 0x7f0b08f24f28>, 'bn_scale': True, 'weight_decay': 0.0, 'kernel_initializer': 'he_uniform', 'first_scale': <function <lambda> at 0x7f0ab4530268>, 'dhw': [32, 32, 32], 'k': 16, 'bottleneck': 4, 'compression': 2, 'first_layer': 32, 'down_structure': [4, 4, 4], 'output_size': 3}\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 32, 32, 32, 1 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "lambda_1 (Lambda)               (None, 32, 32, 32, 1 0           input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_1 (Conv3D)               (None, 32, 32, 32, 3 896         lambda_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 32, 32, 32, 3 128         conv3d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, 32, 32, 32, 3 0           batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_2 (Conv3D)               (None, 32, 32, 32, 6 2048        activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 32, 32, 32, 6 256         conv3d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, 32, 32, 32, 6 0           batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_3 (Conv3D)               (None, 32, 32, 32, 1 27664       activation_2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 32, 32, 32, 4 0           conv3d_3[0][0]                   \n",
      "                                                                 conv3d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 32, 32, 32, 4 192         concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 32, 32, 32, 4 0           batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_4 (Conv3D)               (None, 32, 32, 32, 6 3072        activation_3[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 32, 32, 32, 6 256         conv3d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_4 (Activation)       (None, 32, 32, 32, 6 0           batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_5 (Conv3D)               (None, 32, 32, 32, 1 27664       activation_4[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 32, 32, 32, 6 0           conv3d_5[0][0]                   \n",
      "                                                                 concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 32, 32, 32, 6 256         concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_5 (Activation)       (None, 32, 32, 32, 6 0           batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_6 (Conv3D)               (None, 32, 32, 32, 6 4096        activation_5[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 32, 32, 32, 6 256         conv3d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_6 (Activation)       (None, 32, 32, 32, 6 0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_7 (Conv3D)               (None, 32, 32, 32, 1 27664       activation_6[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 32, 32, 32, 8 0           conv3d_7[0][0]                   \n",
      "                                                                 concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 32, 32, 32, 8 320         concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_7 (Activation)       (None, 32, 32, 32, 8 0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_8 (Conv3D)               (None, 32, 32, 32, 6 5120        activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 32, 32, 32, 6 256         conv3d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_8 (Activation)       (None, 32, 32, 32, 6 0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_9 (Conv3D)               (None, 32, 32, 32, 1 27664       activation_8[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 32, 32, 32, 9 0           conv3d_9[0][0]                   \n",
      "                                                                 concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 32, 32, 32, 9 384         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_9 (Activation)       (None, 32, 32, 32, 9 0           batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_10 (Conv3D)              (None, 32, 32, 32, 4 4656        activation_9[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_1 (AveragePoo (None, 16, 16, 16, 4 0           conv3d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, 16, 16, 16, 4 192         average_pooling3d_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_10 (Activation)      (None, 16, 16, 16, 4 0           batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_11 (Conv3D)              (None, 16, 16, 16, 6 3072        activation_10[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, 16, 16, 16, 6 256         conv3d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_11 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_12 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_11[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_5 (Concatenate)     (None, 16, 16, 16, 6 0           conv3d_12[0][0]                  \n",
      "                                                                 average_pooling3d_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, 16, 16, 16, 6 256         concatenate_5[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_12 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_13 (Conv3D)              (None, 16, 16, 16, 6 4096        activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, 16, 16, 16, 6 256         conv3d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_13 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_14 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_6 (Concatenate)     (None, 16, 16, 16, 8 0           conv3d_14[0][0]                  \n",
      "                                                                 concatenate_5[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, 16, 16, 16, 8 320         concatenate_6[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_14 (Activation)      (None, 16, 16, 16, 8 0           batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_15 (Conv3D)              (None, 16, 16, 16, 6 5120        activation_14[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, 16, 16, 16, 6 256         conv3d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_15 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_16 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_15[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_7 (Concatenate)     (None, 16, 16, 16, 9 0           conv3d_16[0][0]                  \n",
      "                                                                 concatenate_6[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, 16, 16, 16, 9 384         concatenate_7[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_16 (Activation)      (None, 16, 16, 16, 9 0           batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_17 (Conv3D)              (None, 16, 16, 16, 6 6144        activation_16[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, 16, 16, 16, 6 256         conv3d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_17 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_17[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_18 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_17[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_8 (Concatenate)     (None, 16, 16, 16, 1 0           conv3d_18[0][0]                  \n",
      "                                                                 concatenate_7[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_18 (BatchNo (None, 16, 16, 16, 1 448         concatenate_8[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_18 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_18[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_19 (Conv3D)              (None, 16, 16, 16, 5 6328        activation_18[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_2 (AveragePoo (None, 8, 8, 8, 56)  0           conv3d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_19 (BatchNo (None, 8, 8, 8, 56)  224         average_pooling3d_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_19 (Activation)      (None, 8, 8, 8, 56)  0           batch_normalization_19[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_20 (Conv3D)              (None, 8, 8, 8, 64)  3584        activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_20 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_20 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_20[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_21 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_20[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_9 (Concatenate)     (None, 8, 8, 8, 72)  0           conv3d_21[0][0]                  \n",
      "                                                                 average_pooling3d_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_21 (BatchNo (None, 8, 8, 8, 72)  288         concatenate_9[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_21 (Activation)      (None, 8, 8, 8, 72)  0           batch_normalization_21[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_22 (Conv3D)              (None, 8, 8, 8, 64)  4608        activation_21[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_22 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_22 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_22[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_23 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_22[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_10 (Concatenate)    (None, 8, 8, 8, 88)  0           conv3d_23[0][0]                  \n",
      "                                                                 concatenate_9[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_23 (BatchNo (None, 8, 8, 8, 88)  352         concatenate_10[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_23 (Activation)      (None, 8, 8, 8, 88)  0           batch_normalization_23[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_24 (Conv3D)              (None, 8, 8, 8, 64)  5632        activation_23[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_24 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_24[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_24 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_24[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_25 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_11 (Concatenate)    (None, 8, 8, 8, 104) 0           conv3d_25[0][0]                  \n",
      "                                                                 concatenate_10[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_25 (BatchNo (None, 8, 8, 8, 104) 416         concatenate_11[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_25 (Activation)      (None, 8, 8, 8, 104) 0           batch_normalization_25[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_26 (Conv3D)              (None, 8, 8, 8, 64)  6656        activation_25[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_26 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_26[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_26 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_26[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_27 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_26[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_12 (Concatenate)    (None, 8, 8, 8, 120) 0           conv3d_27[0][0]                  \n",
      "                                                                 concatenate_11[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_27 (BatchNo (None, 8, 8, 8, 120) 480         concatenate_12[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_27 (Activation)      (None, 8, 8, 8, 120) 0           batch_normalization_27[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling3d_1 (Glo (None, 120)          0           activation_27[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 3)            363         global_average_pooling3d_1[0][0] \n",
      "==================================================================================================\n",
      "Total params: 405,171\n",
      "Trainable params: 401,315\n",
      "Non-trainable params: 3,856\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "densenet_model = densenet.get_compiled()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model hyper-parameters: {'activation': <function <lambda> at 0x7f0ab4530620>, 'bn_scale': True, 'weight_decay': 0.0, 'kernel_initializer': 'he_uniform', 'first_scale': <function <lambda> at 0x7f0ab4530730>, 'dhw': [32, 32, 32], 'k': 16, 'bottleneck': 4, 'compression': 2, 'first_layer': 32, 'down_structure': [4, 4, 4], 'output_size': 3, 'dropout_rate': None}\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_2 (InputLayer)            (None, 32, 32, 32, 1 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "lambda_2 (Lambda)               (None, 32, 32, 32, 1 0           input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_28 (Conv3D)              (None, 32, 32, 32, 3 896         lambda_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_28 (BatchNo (None, 32, 32, 32, 3 128         conv3d_28[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_28 (Activation)      (None, 32, 32, 32, 3 0           batch_normalization_28[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_29 (Conv3D)              (None, 32, 32, 32, 6 2048        activation_28[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_29 (BatchNo (None, 32, 32, 32, 6 256         conv3d_29[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_29 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_29[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_30 (Conv3D)              (None, 32, 32, 32, 1 27664       activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_13 (Concatenate)    (None, 32, 32, 32, 4 0           conv3d_30[0][0]                  \n",
      "                                                                 conv3d_28[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_30 (BatchNo (None, 32, 32, 32, 4 192         concatenate_13[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_30 (Activation)      (None, 32, 32, 32, 4 0           batch_normalization_30[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_31 (Conv3D)              (None, 32, 32, 32, 6 3072        activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_31 (BatchNo (None, 32, 32, 32, 6 256         conv3d_31[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_31 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_31[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_32 (Conv3D)              (None, 32, 32, 32, 1 27664       activation_31[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_14 (Concatenate)    (None, 32, 32, 32, 6 0           conv3d_32[0][0]                  \n",
      "                                                                 concatenate_13[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_32 (BatchNo (None, 32, 32, 32, 6 256         concatenate_14[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_32 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_32[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_33 (Conv3D)              (None, 32, 32, 32, 6 4096        activation_32[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_33 (BatchNo (None, 32, 32, 32, 6 256         conv3d_33[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_33 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_33[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_34 (Conv3D)              (None, 32, 32, 32, 1 27664       activation_33[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_15 (Concatenate)    (None, 32, 32, 32, 8 0           conv3d_34[0][0]                  \n",
      "                                                                 concatenate_14[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_34 (BatchNo (None, 32, 32, 32, 8 320         concatenate_15[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_34 (Activation)      (None, 32, 32, 32, 8 0           batch_normalization_34[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_35 (Conv3D)              (None, 32, 32, 32, 6 5120        activation_34[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_35 (BatchNo (None, 32, 32, 32, 6 256         conv3d_35[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_35 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_35[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_36 (Conv3D)              (None, 32, 32, 32, 1 27664       activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_16 (Concatenate)    (None, 32, 32, 32, 9 0           conv3d_36[0][0]                  \n",
      "                                                                 concatenate_15[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_36 (BatchNo (None, 32, 32, 32, 9 384         concatenate_16[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_36 (Activation)      (None, 32, 32, 32, 9 0           batch_normalization_36[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_37 (Conv3D)              (None, 32, 32, 32, 4 4656        activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_3 (AveragePoo (None, 16, 16, 16, 4 0           conv3d_37[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_37 (BatchNo (None, 16, 16, 16, 4 192         average_pooling3d_3[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_37 (Activation)      (None, 16, 16, 16, 4 0           batch_normalization_37[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_38 (Conv3D)              (None, 16, 16, 16, 6 3072        activation_37[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_38 (BatchNo (None, 16, 16, 16, 6 256         conv3d_38[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_38 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_38[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_39 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_38[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_17 (Concatenate)    (None, 16, 16, 16, 6 0           conv3d_39[0][0]                  \n",
      "                                                                 average_pooling3d_3[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_39 (BatchNo (None, 16, 16, 16, 6 256         concatenate_17[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_39 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_39[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_40 (Conv3D)              (None, 16, 16, 16, 6 4096        activation_39[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_40 (BatchNo (None, 16, 16, 16, 6 256         conv3d_40[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_40 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_40[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_41 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_40[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_18 (Concatenate)    (None, 16, 16, 16, 8 0           conv3d_41[0][0]                  \n",
      "                                                                 concatenate_17[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_41 (BatchNo (None, 16, 16, 16, 8 320         concatenate_18[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_41 (Activation)      (None, 16, 16, 16, 8 0           batch_normalization_41[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_42 (Conv3D)              (None, 16, 16, 16, 6 5120        activation_41[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_42 (BatchNo (None, 16, 16, 16, 6 256         conv3d_42[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_42 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_42[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_43 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_19 (Concatenate)    (None, 16, 16, 16, 9 0           conv3d_43[0][0]                  \n",
      "                                                                 concatenate_18[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_43 (BatchNo (None, 16, 16, 16, 9 384         concatenate_19[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_43 (Activation)      (None, 16, 16, 16, 9 0           batch_normalization_43[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_44 (Conv3D)              (None, 16, 16, 16, 6 6144        activation_43[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_44 (BatchNo (None, 16, 16, 16, 6 256         conv3d_44[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_44 (Activation)      (None, 16, 16, 16, 6 0           batch_normalization_44[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_45 (Conv3D)              (None, 16, 16, 16, 1 27664       activation_44[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_20 (Concatenate)    (None, 16, 16, 16, 1 0           conv3d_45[0][0]                  \n",
      "                                                                 concatenate_19[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_45 (BatchNo (None, 16, 16, 16, 1 448         concatenate_20[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_45 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_45[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_46 (Conv3D)              (None, 16, 16, 16, 5 6328        activation_45[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_4 (AveragePoo (None, 8, 8, 8, 56)  0           conv3d_46[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_46 (BatchNo (None, 8, 8, 8, 56)  224         average_pooling3d_4[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_46 (Activation)      (None, 8, 8, 8, 56)  0           batch_normalization_46[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_47 (Conv3D)              (None, 8, 8, 8, 64)  3584        activation_46[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_47 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_47[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_47 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_47[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_48 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_47[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_21 (Concatenate)    (None, 8, 8, 8, 72)  0           conv3d_48[0][0]                  \n",
      "                                                                 average_pooling3d_4[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_48 (BatchNo (None, 8, 8, 8, 72)  288         concatenate_21[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_48 (Activation)      (None, 8, 8, 8, 72)  0           batch_normalization_48[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_49 (Conv3D)              (None, 8, 8, 8, 64)  4608        activation_48[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_49 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_49[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_49 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_49[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_50 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_49[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_22 (Concatenate)    (None, 8, 8, 8, 88)  0           conv3d_50[0][0]                  \n",
      "                                                                 concatenate_21[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_50 (BatchNo (None, 8, 8, 8, 88)  352         concatenate_22[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_50 (Activation)      (None, 8, 8, 8, 88)  0           batch_normalization_50[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_51 (Conv3D)              (None, 8, 8, 8, 64)  5632        activation_50[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_51 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_51[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_51 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_51[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_52 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_51[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_23 (Concatenate)    (None, 8, 8, 8, 104) 0           conv3d_52[0][0]                  \n",
      "                                                                 concatenate_22[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_52 (BatchNo (None, 8, 8, 8, 104) 416         concatenate_23[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_52 (Activation)      (None, 8, 8, 8, 104) 0           batch_normalization_52[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_53 (Conv3D)              (None, 8, 8, 8, 64)  6656        activation_52[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_53 (BatchNo (None, 8, 8, 8, 64)  256         conv3d_53[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_53 (Activation)      (None, 8, 8, 8, 64)  0           batch_normalization_53[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_54 (Conv3D)              (None, 8, 8, 8, 16)  27664       activation_53[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_24 (Concatenate)    (None, 8, 8, 8, 120) 0           conv3d_54[0][0]                  \n",
      "                                                                 concatenate_23[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_1 (Conv3DTrans (None, 16, 16, 16, 1 107632      concatenate_24[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_54 (BatchNo (None, 8, 8, 8, 120) 480         concatenate_24[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "add_1 (Add)                     (None, 16, 16, 16, 1 0           concatenate_20[0][0]             \n",
      "                                                                 conv3d_transpose_1[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_54 (Activation)      (None, 8, 8, 8, 120) 0           batch_normalization_54[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_2 (Conv3DTrans (None, 32, 32, 32, 9 86112       add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling3d_2 (Glo (None, 120)          0           activation_54[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_2 (Add)                     (None, 32, 32, 32, 9 0           concatenate_16[0][0]             \n",
      "                                                                 conv3d_transpose_2[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "clf (Dense)                     (None, 3)            363         global_average_pooling3d_2[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "seg (Conv3D)                    (None, 32, 32, 32, 1 97          add_2[0][0]                      \n",
      "==================================================================================================\n",
      "Total params: 599,012\n",
      "Trainable params: 595,156\n",
      "Non-trainable params: 3,856\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "densesharp_model = densesharp.get_compiled()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model hyper-parameters: {'activation': <function <lambda> at 0x7f0ab4530620>, 'bn_scale': True, 'weight_decay': 0.0, 'kernel_initializer': 'he_uniform', 'first_scale': <function <lambda> at 0x7f0ab4530730>, 'dhw': [32, 32, 32], 'k': 32, 'bottleneck': 4, 'compression': 2, 'first_layer': 32, 'down_structure': [4, 4, 4], 'output_size': 3, 'dropout_rate': None}\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_3 (InputLayer)            (None, 32, 32, 32, 1 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "lambda_3 (Lambda)               (None, 32, 32, 32, 1 0           input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_55 (Conv3D)              (None, 32, 32, 32, 3 896         lambda_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_55 (BatchNo (None, 32, 32, 32, 3 128         conv3d_55[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_55 (Activation)      (None, 32, 32, 32, 3 0           batch_normalization_55[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_56 (Conv3D)              (None, 32, 32, 32, 1 4096        activation_55[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_56 (BatchNo (None, 32, 32, 32, 1 512         conv3d_56[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_56 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_56[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_57 (Conv3D)              (None, 32, 32, 32, 3 110624      activation_56[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_25 (Concatenate)    (None, 32, 32, 32, 6 0           conv3d_57[0][0]                  \n",
      "                                                                 conv3d_55[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_57 (BatchNo (None, 32, 32, 32, 6 256         concatenate_25[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_57 (Activation)      (None, 32, 32, 32, 6 0           batch_normalization_57[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_58 (Conv3D)              (None, 32, 32, 32, 1 8192        activation_57[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_58 (BatchNo (None, 32, 32, 32, 1 512         conv3d_58[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_58 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_58[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_59 (Conv3D)              (None, 32, 32, 32, 3 110624      activation_58[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_26 (Concatenate)    (None, 32, 32, 32, 9 0           conv3d_59[0][0]                  \n",
      "                                                                 concatenate_25[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_59 (BatchNo (None, 32, 32, 32, 9 384         concatenate_26[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_59 (Activation)      (None, 32, 32, 32, 9 0           batch_normalization_59[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_60 (Conv3D)              (None, 32, 32, 32, 1 12288       activation_59[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_60 (BatchNo (None, 32, 32, 32, 1 512         conv3d_60[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_60 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_60[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_61 (Conv3D)              (None, 32, 32, 32, 3 110624      activation_60[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_27 (Concatenate)    (None, 32, 32, 32, 1 0           conv3d_61[0][0]                  \n",
      "                                                                 concatenate_26[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_61 (BatchNo (None, 32, 32, 32, 1 512         concatenate_27[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_61 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_61[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_62 (Conv3D)              (None, 32, 32, 32, 1 16384       activation_61[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_62 (BatchNo (None, 32, 32, 32, 1 512         conv3d_62[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_62 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_62[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_63 (Conv3D)              (None, 32, 32, 32, 3 110624      activation_62[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_28 (Concatenate)    (None, 32, 32, 32, 1 0           conv3d_63[0][0]                  \n",
      "                                                                 concatenate_27[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_63 (BatchNo (None, 32, 32, 32, 1 640         concatenate_28[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_63 (Activation)      (None, 32, 32, 32, 1 0           batch_normalization_63[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_64 (Conv3D)              (None, 32, 32, 32, 8 12880       activation_63[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_5 (AveragePoo (None, 16, 16, 16, 8 0           conv3d_64[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_64 (BatchNo (None, 16, 16, 16, 8 320         average_pooling3d_5[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_64 (Activation)      (None, 16, 16, 16, 8 0           batch_normalization_64[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_65 (Conv3D)              (None, 16, 16, 16, 1 10240       activation_64[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_65 (BatchNo (None, 16, 16, 16, 1 512         conv3d_65[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_65 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_65[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_66 (Conv3D)              (None, 16, 16, 16, 3 110624      activation_65[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_29 (Concatenate)    (None, 16, 16, 16, 1 0           conv3d_66[0][0]                  \n",
      "                                                                 average_pooling3d_5[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_66 (BatchNo (None, 16, 16, 16, 1 448         concatenate_29[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_66 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_66[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_67 (Conv3D)              (None, 16, 16, 16, 1 14336       activation_66[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_67 (BatchNo (None, 16, 16, 16, 1 512         conv3d_67[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_67 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_67[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_68 (Conv3D)              (None, 16, 16, 16, 3 110624      activation_67[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_30 (Concatenate)    (None, 16, 16, 16, 1 0           conv3d_68[0][0]                  \n",
      "                                                                 concatenate_29[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_68 (BatchNo (None, 16, 16, 16, 1 576         concatenate_30[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_68 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_68[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_69 (Conv3D)              (None, 16, 16, 16, 1 18432       activation_68[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_69 (BatchNo (None, 16, 16, 16, 1 512         conv3d_69[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_69 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_69[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_70 (Conv3D)              (None, 16, 16, 16, 3 110624      activation_69[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_31 (Concatenate)    (None, 16, 16, 16, 1 0           conv3d_70[0][0]                  \n",
      "                                                                 concatenate_30[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_70 (BatchNo (None, 16, 16, 16, 1 704         concatenate_31[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_70 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_70[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_71 (Conv3D)              (None, 16, 16, 16, 1 22528       activation_70[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_71 (BatchNo (None, 16, 16, 16, 1 512         conv3d_71[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_71 (Activation)      (None, 16, 16, 16, 1 0           batch_normalization_71[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_72 (Conv3D)              (None, 16, 16, 16, 3 110624      activation_71[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_32 (Concatenate)    (None, 16, 16, 16, 2 0           conv3d_72[0][0]                  \n",
      "                                                                 concatenate_31[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_72 (BatchNo (None, 16, 16, 16, 2 832         concatenate_32[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_72 (Activation)      (None, 16, 16, 16, 2 0           batch_normalization_72[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_73 (Conv3D)              (None, 16, 16, 16, 1 21736       activation_72[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling3d_6 (AveragePoo (None, 8, 8, 8, 104) 0           conv3d_73[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_73 (BatchNo (None, 8, 8, 8, 104) 416         average_pooling3d_6[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_73 (Activation)      (None, 8, 8, 8, 104) 0           batch_normalization_73[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_74 (Conv3D)              (None, 8, 8, 8, 128) 13312       activation_73[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_74 (BatchNo (None, 8, 8, 8, 128) 512         conv3d_74[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_74 (Activation)      (None, 8, 8, 8, 128) 0           batch_normalization_74[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_75 (Conv3D)              (None, 8, 8, 8, 32)  110624      activation_74[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_33 (Concatenate)    (None, 8, 8, 8, 136) 0           conv3d_75[0][0]                  \n",
      "                                                                 average_pooling3d_6[0][0]        \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "batch_normalization_75 (BatchNo (None, 8, 8, 8, 136) 544         concatenate_33[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_75 (Activation)      (None, 8, 8, 8, 136) 0           batch_normalization_75[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_76 (Conv3D)              (None, 8, 8, 8, 128) 17408       activation_75[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_76 (BatchNo (None, 8, 8, 8, 128) 512         conv3d_76[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_76 (Activation)      (None, 8, 8, 8, 128) 0           batch_normalization_76[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_77 (Conv3D)              (None, 8, 8, 8, 32)  110624      activation_76[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_34 (Concatenate)    (None, 8, 8, 8, 168) 0           conv3d_77[0][0]                  \n",
      "                                                                 concatenate_33[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_77 (BatchNo (None, 8, 8, 8, 168) 672         concatenate_34[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_77 (Activation)      (None, 8, 8, 8, 168) 0           batch_normalization_77[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_78 (Conv3D)              (None, 8, 8, 8, 128) 21504       activation_77[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_78 (BatchNo (None, 8, 8, 8, 128) 512         conv3d_78[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_78 (Activation)      (None, 8, 8, 8, 128) 0           batch_normalization_78[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_79 (Conv3D)              (None, 8, 8, 8, 32)  110624      activation_78[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_35 (Concatenate)    (None, 8, 8, 8, 200) 0           conv3d_79[0][0]                  \n",
      "                                                                 concatenate_34[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_79 (BatchNo (None, 8, 8, 8, 200) 800         concatenate_35[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_79 (Activation)      (None, 8, 8, 8, 200) 0           batch_normalization_79[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_80 (Conv3D)              (None, 8, 8, 8, 128) 25600       activation_79[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_80 (BatchNo (None, 8, 8, 8, 128) 512         conv3d_80[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_80 (Activation)      (None, 8, 8, 8, 128) 0           batch_normalization_80[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_81 (Conv3D)              (None, 8, 8, 8, 32)  110624      activation_80[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_36 (Concatenate)    (None, 8, 8, 8, 232) 0           conv3d_81[0][0]                  \n",
      "                                                                 concatenate_35[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_3 (Conv3DTrans (None, 16, 16, 16, 2 386256      concatenate_36[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_81 (BatchNo (None, 8, 8, 8, 232) 928         concatenate_36[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "add_3 (Add)                     (None, 16, 16, 16, 2 0           concatenate_32[0][0]             \n",
      "                                                                 conv3d_transpose_3[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_81 (Activation)      (None, 8, 8, 8, 232) 0           batch_normalization_81[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_4 (Conv3DTrans (None, 32, 32, 32, 1 266400      add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling3d_3 (Glo (None, 232)          0           activation_81[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_4 (Add)                     (None, 32, 32, 32, 1 0           concatenate_28[0][0]             \n",
      "                                                                 conv3d_transpose_4[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "clf (Dense)                     (None, 3)            699         global_average_pooling3d_3[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "seg (Conv3D)                    (None, 32, 32, 32, 1 161         add_4[0][0]                      \n",
      "==================================================================================================\n",
      "Total params: 2,215,140\n",
      "Trainable params: 2,207,988\n",
      "Non-trainable params: 7,152\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "densesharp_model = densesharp.get_compiled(k=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}