nvidia-AI.summerschool17-no.../teaching-lw/misc/temp-speech-worksheet.ipynb

737 lines
1.5 MiB
Text
Raw Permalink Normal View History

2017-09-22 17:26:46 +00:00
{
"metadata": {
"name": "speech-worksheet-again"
},
"nbformat": 2,
"worksheets": [
{
"cells": [
{
"cell_type": "code",
"collapsed": false,
"input": [
"!ls ~/Downloads"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"MHZ-Infobroschuere-Plissee-Vorhaenge-2012 (1).pdf",
"MHZ-Infobroschuere-Plissee-Vorhaenge-2012.pdf",
"audio.tgz",
"fenster_tueren.pdf",
"google-talkplugin_current_amd64.deb"
]
}
],
"prompt_number": 2
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!ls msak0_306.*"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"msak0_306.ema msak0_306.epg msak0_306.lab msak0_306.lar msak0_306.wav"
]
}
],
"prompt_number": 4
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!file msak0_306.wav"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"msak0_306.wav: NIST SPHERE file"
]
}
],
"prompt_number": 5
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!sed '/^$/q' msak0_306.wav"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"NIST_1A",
" 1024",
"channel_count -i 1",
"sample_count -i 59484",
"sample_rate -i 16000",
"sample_coding -s3 pcm",
"sample_n_bytes -i 2",
"sample_byte_format -s2 01",
"microphone -s21 audio-technica ATM10a",
"recording_site -s33 Queen Margaret University College",
"database_id -s11 MOCHA-TIMIT",
"database_version -s3 1.0",
"recording_environment -s26 sound-damped studio (4x4m)",
"speaking_mode -s4 read",
"dialect -s16 Northern England",
"sex -s4 male",
"recording_date -s12 21-Sept-1999",
"licence -s32 Non-commercial (See Licence.txt)",
"sample_checksum -i 49246",
"end_head"
]
}
],
"prompt_number": 7
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!play msak0_306.wav"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"msak0_306.wav:",
"",
" File Size: 120k Bit Rate: 258k",
" Encoding: Signed PCM ",
" Channels: 1 @ 16-bit ",
"Samplerate: 16000Hz ",
"Replaygain: off ",
" Duration: 00:00:03.72 ",
"",
"",
"In:0.00% 00:00:00.00 [00:00:03.72] Out:0 [ | ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:68.9% 00:00:02.56 [00:00:01.16] Out:41.0k [ ===|=== ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:82.6% 00:00:03.07 [00:00:00.65] Out:49.2k [ -==|==- ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:96.4% 00:00:03.58 [00:00:00.13] Out:57.3k [ | ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:100% 00:00:03.72 [00:00:00.00] Out:59.5k [ | ] Clip:0 ",
"In:100% 00:00:03.72 [00:00:00.00] Out:59.5k [ | ] Clip:0 ",
"Done."
]
}
],
"prompt_number": 12
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"!sox msak0_306.wav temp.raw"
],
"language": "python",
"outputs": [],
"prompt_number": 13
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!play temp.raw"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"play FAIL formats: bad input format for file `temp.raw': sampling rate was not specified"
]
}
],
"prompt_number": 14
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!play -s2L -r 16000 temp.raw"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"temp.raw:",
"",
" File Size: 119k Bit Rate: 256k",
" Encoding: Signed PCM ",
" Channels: 1 @ 16-bit ",
"Samplerate: 16000Hz ",
"Replaygain: off ",
" Duration: 00:00:03.72 ",
"",
"",
"In:0.00% 00:00:00.00 [00:00:03.72] Out:0 [ | ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:68.9% 00:00:02.56 [00:00:01.16] Out:41.0k [ ===|=== ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:82.6% 00:00:03.07 [00:00:00.65] Out:49.2k [ -==|==- ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:96.4% 00:00:03.58 [00:00:00.13] Out:57.3k [ | ] Clip:0 "
]
},
{
"output_type": "stream",
"stream": "stdout",
"text": [
"",
"In:100% 00:00:03.72 [00:00:00.00] Out:59.5k [ | ] Clip:0 ",
"In:100% 00:00:03.72 [00:00:00.00] Out:59.5k [ | ] Clip:0 ",
"Done."
]
}
],
"prompt_number": 17
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"import numpy",
"from pylab import *"
],
"language": "python",
"outputs": [],
"prompt_number": 18
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"data = numpy.fromfile(\"temp.raw\",\"int16\")"
],
"language": "python",
"outputs": [],
"prompt_number": 41
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"data.shape"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 42,
"text": [
"(59484,)"
]
}
],
"prompt_number": 42
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"plot(data[10000:15000])"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 43,
"text": [
"[<matplotlib.lines.Line2D at 0x5265b90>]"
]
},
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAD9CAYAAABA8iukAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ny0vikhiMaFBAMS1IZIOLNEwW
}
],
"prompt_number": 43
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"plot(data)"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 44,
"text": [
"[<matplotlib.lines.Line2D at 0x542b2d0>]"
]
},
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAD9CAYAAACSoiH8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLFmCsrIyREdHY9euXViyZAkA
}
],
"prompt_number": 44
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"!cat msak0_306.lab"
],
"language": "python",
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"0 0.2 sil",
"0.2 0.49 breath",
"0.49 0.69 sil",
"0.69 0.76 w",
"0.76 0.89 ii",
"0.89 0.94 l",
"0.94 1.04 ai",
"1.04 1.07 k",
"1.07 1.17 b",
"1.17 1.22 l",
"1.22 1.33 uu",
"1.33 1.45 ch",
"1.45 1.59 ii",
"1.59 1.68 z",
"1.68 1.74 b",
"1.74 1.82 @",
"1.82 1.85 t",
"1.85 1.92 v",
"1.92 2 i",
"2 2.04 k",
"2.04 2.12 t",
"2.12 2.16 @",
"2.16 2.21 l",
"2.21 2.31 ai",
"2.31 2.34 k",
"2.34 2.4 s",
"2.4 2.49 b",
"2.49 2.58 r",
"2.58 2.74 ii",
"2.74 3.7 sil"
]
}
],
"prompt_number": 45
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"!rm _out.raw"
],
"language": "python",
"outputs": [],
"prompt_number": 39
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"import os",
"def play(data,pad=8000):",
" data = r_[zeros(pad,data.dtype),data,zeros(pad,data.dtype)]",
" data.tofile(\"_out.raw\")",
" os.system(\"play -s2L -r 16000 _out.raw\")",
"play(data)"
],
"language": "python",
"outputs": [],
"prompt_number": 46
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"play(array(list(data[int(1.33*16000):int(1.68*16000)])*10))"
],
"language": "python",
"outputs": [],
"prompt_number": 52
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def read_lab(fname):",
" result = []",
" with open(fname) as stream:",
" for line in stream.readlines():",
" line = line[:-1]",
" start,stop,label = line.split(\" \",2)",
" result.append((float(start),float(stop),label))",
" return result",
"read_lab(\"msak0_306.lab\")"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 53,
"text": [
"[(0.0, 0.2, 'sil'),",
" (0.2, 0.49, 'breath'),",
" (0.49, 0.69, 'sil'),",
" (0.69, 0.76, 'w'),",
" (0.76, 0.89, 'ii'),",
" (0.89, 0.94, 'l'),",
" (0.94, 1.04, 'ai'),",
" (1.04, 1.07, 'k'),",
" (1.07, 1.17, 'b'),",
" (1.17, 1.22, 'l'),",
" (1.22, 1.33, 'uu'),",
" (1.33, 1.45, 'ch'),",
" (1.45, 1.59, 'ii'),",
" (1.59, 1.68, 'z'),",
" (1.68, 1.74, 'b'),",
" (1.74, 1.82, '@'),",
" (1.82, 1.85, 't'),",
" (1.85, 1.92, 'v'),",
" (1.92, 2.0, 'i'),",
" (2.0, 2.04, 'k'),",
" (2.04, 2.12, 't'),",
" (2.12, 2.16, '@'),",
" (2.16, 2.21, 'l'),",
" (2.21, 2.31, 'ai'),",
" (2.31, 2.34, 'k'),",
" (2.34, 2.4, 's'),",
" (2.4, 2.49, 'b'),",
" (2.49, 2.58, 'r'),",
" (2.58, 2.74, 'ii'),",
" (2.74, 3.7, 'sil')]"
]
}
],
"prompt_number": 53
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"import os",
"class Wave:",
" pass",
"def read_data(fname):",
" os.system(\"sox %s _temp.raw\"%fname)",
" data = numpy.fromfile(\"_temp.raw\",'int16')",
" wave = Wave() ",
" base,_ = os.path.splitext(fname)",
" wave.labels = read_lab(base+\".lab\")",
" wave.data = data",
" wave.rate = 16000.0",
" return wave",
"wave = read_data(\"msak0_306.wav\")"
],
"language": "python",
"outputs": [],
"prompt_number": 55
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"def phonemes(wave):",
" for start,stop,label in wave.labels:",
" start = int(wave.rate*start)",
" stop = int(wave.rate*stop)",
" #print start,stop,len(wave.data)",
" clip = wave.data[start:stop]",
" yield clip,label",
"clips = list(phonemes(wave))"
],
"language": "python",
"outputs": [],
"prompt_number": 56
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"clips[0]"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 57,
"text": [
"(array([ 470, 827, -1059, ..., 72, 87, 94], dtype=int16), 'sil')"
]
}
],
"prompt_number": 57
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"import glob",
"samples = []",
"for fname in sorted(glob.glob(\"msak*.wav\")):",
" wave = read_data(fname)",
" for clip,cls in phonemes(wave):",
" samples.append((clip,cls))"
],
"language": "python",
"outputs": [],
"prompt_number": 58
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from collections import Counter",
"classes = [cls for clip,cls in samples]",
"Counter(classes)"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 59,
"text": [
"Counter({'@': 1388, 'sil': 1380, 'i': 938, 't': 889, 'n': 840, 's': 698, 'l': 666, 'r': 615, 'k': 541, 'd': 536, 'z': 495, 'breath': 460, 'm': 420, 'p': 374, 'dh': 331, 'b': 309, 'ii': 305, 'e': 297, 'f': 264, 'uu': 263, 'ai': 251, 'ei': 249, 'w': 246, 'o': 230, 'a': 228, 'v': 227, 'oo': 203, 'ou': 202, 'y': 197, 'g': 195, 'uh': 186, 'iy': 167, 'h': 154, 'sh': 147, 'ng': 140, 'jh': 138, 'aa': 106, 'ch': 99, '@@': 90, 'ow': 87, 'th': 73, 'u': 54, 'oi': 44, 'eir': 33, 'i@': 28, 'zh': 17})"
]
}
],
"prompt_number": 59
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"wave = read_data(\"msak0_306.wav\")",
"figsize(12,4)",
"_=specgram(wave.data,Fs=16000,cmap=cm.GnBu)",
"play(wave.data)"
],
"language": "python",
"outputs": [
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAs8AAAD9CAYAAABDX5o4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nns4dMRSKVmbteXdMpGDt/BpD
}
],
"prompt_number": 65
},
{
"cell_type": "code",
"collapsed": true,
"input": [],
"language": "python",
"outputs": []
},
{
"cell_type": "code",
"collapsed": true,
"input": [
"data = numpy.fromfile(\"msak0_306.raw\",\"int16\")"
],
"language": "python",
"outputs": [],
"prompt_number": 66
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"figsize(12,6)",
"_=specgram(data,Fs=16000)"
],
"language": "python",
"outputs": [
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAs8AAAFtCAYAAAD4TofPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niHp1O0CkgnzWxzYRWrfiZ2jo
}
],
"prompt_number": 78
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"len(data)"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 74,
"text": [
"59484"
]
}
],
"prompt_number": 74
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"figsize(6,6)",
"n = 256",
"_=specgram(data[10000:20000],NFFT=n,noverlap=n/2)"
],
"language": "python",
"outputs": [
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAFtCAYAAADmscqNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuPmVfw4GjC2GOrdlzwUcOasp
}
],
"prompt_number": 82
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"256.0/16000"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 83,
"text": [
"0.016"
]
}
],
"prompt_number": 83
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def windowed_spec(data):",
" n = len(data)",
" w = (arange(n)-n/2.0)**2",
" w /= amax(w)",
" w = 1-w",
" return log(abs(fft(w*data)))[:n//2]",
"plot(windowed_spec(data[:256]))"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 84,
"text": [
"[<matplotlib.lines.Line2D at 0x9d3e990>]"
]
},
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAFtCAYAAAAXupEAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYejfvz+uuOKKsPdJxgwdYC+6
}
],
"prompt_number": 84
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"figsize(12,6)",
"",
"spec = array([windowed_spec(wave.data[i:i+256]) for i in range(0,len(wave.data)-256,128)])",
"imshow(spec.T,origin='lower')"
],
"language": "python",
"outputs": [
{
"output_type": "pyout",
"prompt_number": 86,
"text": [
"<matplotlib.image.AxesImage at 0xad59750>"
]
},
{
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAsEAAADTCAYAAABgMM7NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nab3Dg+Ut6lRRqQCBpceE2ObE
}
],
"prompt_number": 86
},
{
"cell_type": "code",
"collapsed": true,
"input": [],
"language": "python",
"outputs": []
}
]
}
]
}