Show_program.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import librosa\n",
    "import keras\n",
    "import keras.layers as L\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import pickle\n",
    "import os\n",
    "import random\n",
    "import bisect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "time_step = 5 #seconds\n",
    "hop_len = 512\n",
    "sr = 22050\n",
    "model=keras.models.Sequential()\n",
    "model.add(L.InputLayer(input_shape=(int(time_step*sr/(hop_len)),128)))#128,215)))\n",
    "model.add(L.Conv1D(256,6, padding='same'))\n",
    "model.add(L.Conv1D(256,3, padding='same'))\n",
    "model.add(L.GlobalMaxPool1D())\n",
    "model.add(L.Dense(512,activation='relu'))\n",
    "model.add(L.Dense(256, activation='relu'))\n",
    "model.add(L.Dense(128, activation='sigmoid'))\n",
    "#model.add(L.GlobalMaxPool1D())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='SGD',loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model.load_weights(\"/home/jheuristic/data/weights/model1.wght\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def generate_for_test(spectrogram,startTime,time_step=5,hop_len = 512,sr = 22050):\n",
    "    startSpectr= int(startTime*sr/(hop_len)) #int(startMid/(k*hop_len))\n",
    "    stopSpectr=int(startSpectr+time_step*sr/(hop_len))\n",
    "    \n",
    "    out1 = spectrogram[:,startSpectr:stopSpectr].transpose().reshape(1,int(time_step*sr/(hop_len)),128)\n",
    "    return out1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def generate_for_test2(spectrogram,startTime,time_step=5,hop_len = 512,sr = 22050):\n",
    "    startSpectr= int(startTime*sr/(hop_len)) #int(startMid/(k*hop_len))\n",
    "    stopSpectr=int(startSpectr+time_step*sr/(hop_len))\n",
    "    \n",
    "    out1 = spectrogram[:,startSpectr:stopSpectr].transpose().reshape(1,1,int(time_step*sr/(hop_len)),128)\n",
    "    return out1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "hjk,sr=librosa.load('/home/jheuristic/data/Untitled_819855_1.wav')\n",
    "spectrogram = librosa.feature.melspectrogram(hjk,sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "trac_len = len(spectrogram[0])*hop_len\n",
    "notes = np.zeros((int(trac_len/(sr*time_step)),128))\n",
    "#trac_len = len(spectrogram[0])*hop_len\n",
    "#notes = np.zeros((int(trac_len/(sr*time_step)),128))\n",
    "for i in range(int(trac_len/(sr*time_step))):\n",
    "    notes[i]=model.predict(generate_for_test(spectrogram,i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f21b7841390>"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUkAAAHVCAYAAAB8LgmQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npIhIARVJEZECKpIiIgVUJEVECvwDPs6QcLtEJkkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f21b7b63588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(24,8))\n",
    "plt.imshow(notes.transpose())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f21b75dec88>"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArwAAAHVCAYAAAATqShMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f21b795ef28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(24,8))\n",
    "plt.imshow(notes > 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "spectrogram = spectrogram.transpose()\n",
    "spectrogram  = np.concatenate((spectrogram,spectrogram,spectrogram,spectrogram))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(128, 7488)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spectrogram = spectrogram.transpose()\n",
    "spectrogram.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "spectrogram=np.random.random(size=(128, 7488))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "spectrogram-=1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "spectrogram[0][0] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test,sr=librosa.load('/home/jheuristic/data/Not_recognizeable.wav')\n",
    "spectrogram = librosa.feature.melspectrogram(test,sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f21bc3b29b0>"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABWcAAAC+CAYAAABZJ71cAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f21bc3cb630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(24,8))\n",
    "plt.imshow(spectrogram)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-5.0"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.min(spectrogram)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "time_step = 5 #seconds\n",
    "hop_len = 512\n",
    "sr = 22050\n",
    "model2=keras.models.Sequential()\n",
    "model2.add(L.InputLayer(input_shape=(1,int(time_step*sr/(hop_len)),128)))#128,215)))\n",
    "model2.add(L.Conv2D(256,(6,6), padding='same'))\n",
    "model2.add(L.Conv2D(256,(3,3), padding='same'))\n",
    "model2.add(L.GlobalMaxPool2D())\n",
    "model2.add(L.Dense(512,activation='relu'))\n",
    "model2.add(L.Dense(256, activation='relu'))\n",
    "model2.add(L.Dense(128, activation='sigmoid'))\n",
    "#model.add(L.GlobalMaxPool1D())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model2.compile(optimizer='SGD',loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model2.load_weights(\"/home/jheuristic/data/weights/model2.wght\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "trac_len = len(spectrogram[0])*hop_len\n",
    "notes = np.zeros((int(trac_len/(sr*time_step)),128))\n",
    "#trac_len = len(spectrogram[0])*hop_len\n",
    "#notes = np.zeros((int(trac_len/(sr*time_step)),128))\n",
    "for i in range(int(trac_len/(sr*time_step))):\n",
    "    notes[i]=model2.predict(generate_for_test2(spectrogram,i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f21b757fd30>"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUkAAAHVCAYAAAB8LgmQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f21b75f4860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(24,8))\n",
    "plt.imshow(notes.transpose())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f21b74a65c0>"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUkAAAHVCAYAAAB8LgmQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f21b7523518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(24,8))\n",
    "plt.imshow(notes.transpose() > 0.5)"
   ]
  }
 ],
 "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}