Merge branch 'IT17100908' into 'QA_RELEASE'

It17100908

See merge request !28

LectureSummarizingApp/noise.py

 import librosa
 from pysndfx import AudioEffectsChain
-import numpy as np
-import math
 import python_speech_features
-import scipy as sp
-from scipy import signal
-import soundfile
+
 
 
 def read_file(file_name):
@@ -14,31 +10,18 @@ def read_file(file_name):
     sample_path = sample_directory + sample_file
 
     # generating audio time series and a sampling rate (int)
-    y, sr = librosa.load(sample_path)
-
-    return y, sr
-
+    a, sr = librosa.load(sample_path)
 
-# '''CENTROID'''
-#
-# def reduce_noise_centroid_s(y, sr):
-#
-#     cent = librosa.feature.spectral_centroid(y=y, sr=sr)
-#     threshold_h = np.max(cent)
-#     threshold_l = np.min(cent)
-#     less_noise = AudioEffectsChain().lowshelf(gain=-12.0, frequency=threshold_l, slope=0.5).highshelf(gain=-12.0, frequency=threshold_h, slope=0.5).limiter(gain=6.0)
-#     y_cleaned = less_noise(y)
-#     return y_cleaned
-#jijoijij
+    return a, sr
 
 
 '''MFCC'''
 
-def mffc_highshelf(y, sr):
+def mffc_highshelf(a, sr):
 
 
-    mfcc = python_speech_features.base.mfcc(y)
-    mfcc = python_speech_features.base.logfbank(y)
+    mfcc = python_speech_features.base.mfcc(a)
+    mfcc = python_speech_features.base.logfbank(a)
     mfcc = python_speech_features.base.lifter(mfcc)
 
     sum_of_squares = []
@@ -56,22 +39,22 @@ def mffc_highshelf(y, sr):
     min_hz = min(hz)
 
     speech_booster = AudioEffectsChain().highshelf(frequency=min_hz*(-1)*1.2, gain=-12.0, slope=0.6).limiter(gain=8.0)
-    y_speach_boosted = speech_booster(y)
+    a_speach_boosted = speech_booster(a)
 
-    return (y_speach_boosted)
+    return (a_speach_boosted)
 
-def mfcc_lowshelf(y, sr):
+def mfcc_lowshelf(a, sr):
 
-    mfcc = python_speech_features.base.mfcc(y)
-    mfcc = python_speech_features.base.logfbank(y)
+    mfcc = python_speech_features.base.mfcc(a)
+    mfcc = python_speech_features.base.logfbank(a)
     mfcc = python_speech_features.base.lifter(mfcc)
 
     sum_of_squares = []
     index = -1
-    for r in mfcc:
+    for x in mfcc:
         sum_of_squares.append(0)
         index = index + 1
-        for n in r:
+        for n in x:
             sum_of_squares[index] = sum_of_squares[index] + n**2
 
     strongest_frame = sum_of_squares.index(max(sum_of_squares))
@@ -81,49 +64,49 @@ def mfcc_lowshelf(y, sr):
     min_hz = min(hz)
 
     speech_booster = AudioEffectsChain().lowshelf(frequency=min_hz*(-1), gain=12.0, slope=0.5)
-    y_speach_boosted = speech_booster(y)
+    a_speach_boosted = speech_booster(a)
 
-    return (y_speach_boosted)
+    return (a_speach_boosted)
 
 
 def trim_silence(y):
-    y_trimmed, index = librosa.effects.trim(y, top_db=20, frame_length=2, hop_length=500)
-    trimmed_length = librosa.get_duration(y) - librosa.get_duration(y_trimmed)
+    a_trimmed, index = librosa.effects.trim(y, top_db=20, frame_length=2, hop_length=500)
+    trimmed_length = librosa.get_duration(y) - librosa.get_duration(a_trimmed)
 
-    return y_trimmed, trimmed_length
+    return a_trimmed, trimmed_length
 
 
 def enhance(y):
     apply_audio_effects = AudioEffectsChain().lowshelf(gain=10.0, frequency=260, slope=0.1).reverb(reverberance=25, hf_damping=5, room_scale=5, stereo_depth=50, pre_delay=20, wet_gain=0, wet_only=False)#.normalize()
-    y_enhanced = apply_audio_effects(y)
+    a_enhanced = apply_audio_effects(y)
 
-    return y_enhanced
+    return a_enhanced
 
-def output_file(destination ,filename, y, sr, ext=""):
+def output_file(destination ,filename, a, sr, ext=""):
     destination = destination + filename[:-4] + ext + '.wav'
-    librosa.output.write_wav(destination, y, sr)
+    librosa.output.write_wav(destination, a, sr)
 
 
 lectures = ['Lecture01.wav']
 
 for s in lectures:
     filename = s
-    y, sr = read_file(filename)
+    a, sr = read_file(filename)
 
 
-    # y_reduced_centroid_s = reduce_noise_centroid_s(y, sr)
-    y_reduced_mfcc_lowshelf = mfcc_lowshelf(y, sr)
-    y_reduced_mfcc_highshelf = mffc_highshelf(y, sr)
+    # a_reduced_centroid_s = reduce_noise_centroid_s(a, sr)
+    a_reduced_mfcc_lowshelf = mfcc_lowshelf(a, sr)
+    a_reduced_mfcc_highshelf = mffc_highshelf(a, sr)
 
 
     # trimming silences
-    # y_reduced_centroid_s, time_trimmed = trim_silence(y_reduced_centroid_s)
-    y_reduced_mfcc_up, time_trimmed = trim_silence(mfcc_lowshelf)
-    y_reduced_mfcc_down, time_trimmed = trim_silence(mffc_highshelf)
+    # a_reduced_centroid_s, time_trimmed = trim_silence(a_reduced_centroid_s)
+    a_reduced_mfcc_up, time_trimmed = trim_silence(mfcc_lowshelf)
+    a_reduced_mfcc_down, time_trimmed = trim_silence(mffc_highshelf)
 
 
 
     # output_file('lectures_trimmed_noise_reduced/' ,filename, y_reduced_centroid_s, sr, '_ctr_s')
-    output_file('lectures_trimmed_noise_reduced/' ,filename, y_reduced_mfcc_up, sr, '_mfcc_up')
-    # output_file('lectures_trimmed_noise_reduced/' ,filename, y_reduced_mfcc_down, sr, '_mfcc_down')
-    # output_file('lectures_trimmed_noise_reduced/' ,filename, y, sr, '_org')
+    output_file('lectures_trimmed_noise_reduced/' ,filename, a_reduced_mfcc_up, sr, '_mfcc_up')
+    # output_file('lectures_trimmed_noise_reduced/' ,filename, a_reduced_mfcc_down, sr, '_mfcc_down')
+    # output_file('lectures_trimmed_noise_reduced/' ,filename, a, sr, '_org')

LectureSummarizingApp/noiseRemove.py

+import numpy as nump
+import scipy as sip
+from scipy.io.wavfile import read
+from scipy.io.wavfile import write
+from scipy import signal
+import matplotlib.pyplot as mplt
+#get_ipython().magic('matplotlib inline')
+
+(Frequency, array) = read('lectures/Lecture01.wav')
+
+len(array)
+
+mplt.plot(array)
+mplt.title('Original Signal Spectrum')
+mplt.xlabel('Frequency(Hz)')
+mplt.ylabel('Amplitude')
+
+fourierTransformation = sip.fft(array)
+
+scale = sip.linspace(0, Frequency, len(array))
+
+mplt.stem(scale[0:5000], nump.abs(fourierTransformation[0:5000]), 'r')
+mplt.title('Signal spectrum after FFT')
+mplt.xlabel('Frequency(Hz)')
+mplt.ylabel('Amplitude')
+
+
+guassianNoise = nump.random.rand(len(fourierTransformation))
+
+
+NewSound = guassianNoise + array
+
+write("New-Sound-Added-With-Guassian-Noise.wav", Frequency, NewSound)
+
+u,v = signal.butter(5, 1000/(Frequency/2), btype='highpass')
+
+filteredSignal = signal.lfilter(u,v,NewSound)
+
+# plotting the signal.
+mplt.plot(filteredSignal)
+mplt.title('Highpass Filter')
+mplt.xlabel('Frequency(Hz)')
+mplt.ylabel('Amplitude')
+
+# ButterWorth low-filter
+x,y = signal.butter(5, 380/(Frequency/2), btype='lowpass')
+
+# Applying the filter to the signal
+newFilteredSignal = signal.lfilter(x,y,filteredSignal)
+
+# plotting the signal.
+mplt.plot(newFilteredSignal)
+mplt.title('Lowpass Filter')
+mplt.xlabel('Frequency(Hz)')
+mplt.ylabel('Amplitude')
+
+write("removed.wav", Frequency, nump.int16(newFilteredSignal/nump.max(nump.abs(newFilteredSignal)) * 32767))
\ No newline at end of file