def compare_ir():
fs = 16e3
# impulse
x_len = np.int16(fs)
x = np.zeros(x_len)
x[1] = 1
gtf_obj = gtf_proposed(fs, cf_low=100, cf_high=2000, n_band=4)
irs = gtf_obj.filter(x)
fig = gtf_obj.plot_ir_spec(irs[:, :1000])
savefig(fig, 'proposed.png')
coefs = gtf_ref.make_erb_filters(fs, gtf_obj.cfs)
irs_ref = gtf_ref.erb_filterbank(x, coefs)
fig = gtf_obj.plot_ir_spec(irs[:, :1000])
savefig(fig, "ref.png")
irs_eq = gtf_obj.get_ir_equation()
fig, ax = plt.subplots(1, 2, sharex=True, sharey=True, tight_layout=True)
ax[0].plot(irs[3] / np.max(irs[3]), label='Todd')
ax[0].plot(irs_eq[3] / np.max(irs_eq[3]), label='Equation')
ax[0].legend()
ax[0].set_xlim([0, 200])
ax[1].plot(irs_ref[3] / np.max(irs_ref[3]), label='Detly')
ax[1].plot(irs_eq[3] / np.max(irs_eq[3]), label='Equation')
ax[1].legend()
ax[1].set_xlim([0, 200])
savefig(fig, 'images/validate/compare.png')
def validate():
fs = 16e3
# impulse
x_len = np.int16(fs)
x = np.zeros(x_len)
x[1] = 1
gtf_obj = gtf_proposed(fs, cf_low=100, cf_high=2000, n_band=4)
irs = gtf_obj.filter(x)
# fig1 = gtf_obj.plot_ir_spec(irs1[:, :1000])
# savefig(fig1, 'proposed.png')
coefs = gtf_reference.make_erb_filters(fs, gtf_obj.cfs)
irs_ref = gtf_reference.erb_filterbank(x, coefs)
# fig2 = gtf_obj.plot_ir_spec(irs2[:, :1000])
# savefig(fig2, "reference.png")
irs_eq = gtf_obj.get_ir_equation()
fig, ax = plt.subplots(1, 2, sharex=True, sharey=True, tight_layout=True)
ax[0].plot(irs[3] / np.max(irs[3]), label='todd')
ax[0].plot(irs_eq[3] / np.max(irs_eq[3]), label='eq')
ax[0].legend()
ax[0].set_xlim([0, 200])
ax[1].plot(irs_ref[3] / np.max(irs_ref[3]), label='detly')
ax[1].plot(irs_eq[3] / np.max(irs_eq[3]), label='eq')
ax[1].legend()
ax[1].set_xlim([0, 200])
savefig(fig, 'compare.png')
def gtgram_xe(wave, fs, channels, f_min, f_max):
""" Calculate the intermediate ERB filterbank processed matrix """
cfs = centre_freqs(fs, channels, f_min, f_max)
fcoefs = np.flipud(gf.make_erb_filters(fs, cfs))
xf = gf.erb_filterbank(wave, fcoefs)
xe = np.power(xf, 2)
return xe
def process_signal(self, in_sig, verbose=False):
'''
Where all the actual signal processing is done -- actually, where all
the signal processing functions and methods are called. Results are all
stored in "self.chunks," which is a collection of ti
'''
self.in_sig = in_sig
print("len: ", len(self.in_sig))
# fdl_out_chunks = []
# agc_out_chunks = []
if self.filt_type == 'bessel':
filted_channels = np.zeros((self.num_chan, len(self.in_sig)))
print(filted_channels.shape)
for k in range(self.num_chan):
if verbose:
print("Processing channel %d/%d" % (k + 1, self.num_chan))
filted_channels[k] = dsp.filtfilt(self.b[k], self.a[k], in_sig)
filted = filted_channels[k]
f0s, idx_chunks, out_chunks, num_chunks = self.fdl[
k].process_data(filted)
for j in range(num_chunks):
if len(out_chunks[j]) < np.floor(
0.03 / self.dt): # dur > 30 ms
continue
# fdl_out_chunks.append(out_chunks[j])
out_chunks[j] = scfbutils.agc(out_chunks[j], 0.1, 0.25)
out_chunks[j] = scfbutils.agc(out_chunks[j], 0.001, 0.25)
# agc_out_chunks.append(out_chunks[j])
freq_est = scfbutils.pll(out_chunks[j], f0s[j], self.f_s)
assert len(freq_est) == len(idx_chunks[j])
self.chunks.append((idx_chunks[j], freq_est))
elif self.filt_type == 'gammatone':
filted_channels = gtf.erb_filterbank(self.in_sig, self.erb_coefs)
# filted_channels = np.flipud(filted_channels)
for k, filted in enumerate(filted_channels):
if verbose:
print("Processing channel %d/%d" % (k + 1, self.num_chan))
f0s, idx_chunks, out_chunks, num_chunks = self.fdl[
k].process_data(filted)
self.out_chunks = out_chunks
for j in range(num_chunks):
if len(out_chunks[j]) < np.floor(
0.03 / self.dt): # dur > 30 ms
continue
# fdl_out_chunks.append(out_chunks[j])
out_chunks[j], amps1 = scfbutils.agc(
out_chunks[j], 0.1, 0.25)
out_chunks[j], amps2 = scfbutils.agc(
out_chunks[j], 0.001, 0.25)
# agc_out_chunks.append(out_chunks[j])
freq_est = scfbutils.pll(out_chunks[j], f0s[j], self.f_s)
assert len(freq_est) == len(idx_chunks[j])
self.chunks.append((idx_chunks[j], freq_est))
else:
print("Failure: given filter type unavailable.")
return 0
self.filtered_channels = filted_channels
self.processed = True
return self.chunks # final goal, next one is for debugging
def srmr(x, fs, n_cochlear_filters=23, low_freq=125, min_cf=4, max_cf=128, fast=True, norm=False):
wLengthS = .256
wIncS = .064
# Computing gammatone envelopes
if fast:
mfs = 400.0
gt_env = fft_gtgram(x, fs, 0.010, 0.0025, n_cochlear_filters, low_freq)
else:
cfs = centre_freqs(fs, n_cochlear_filters, low_freq)
fcoefs = make_erb_filters(fs, cfs)
gt_env = np.abs(hilbert(erb_filterbank(x, fcoefs)))
mfs = fs
wLength = int(np.ceil(wLengthS*mfs))
wInc = int(np.ceil(wIncS*mfs))
# Computing modulation filterbank with Q = 2 and 8 channels
mod_filter_cfs = compute_modulation_cfs(min_cf, max_cf, 8)
MF = modulation_filterbank(mod_filter_cfs, mfs, 2)
n_frames = int(1 + (gt_env.shape[1] - wLength)//wInc)
w = hamming(wLength+1)[:-1] # window is periodic, not symmetric
energy = np.zeros((n_cochlear_filters, 8, n_frames))
for i, ac_ch in enumerate(gt_env):
mod_out = modfilt(MF, ac_ch)
for j, mod_ch in enumerate(mod_out):
mod_out_frame = segment_axis(mod_ch, wLength, overlap=wLength-wInc, end='pad')
energy[i,j,:] = np.sum((w*mod_out_frame[:n_frames])**2, axis=1)
if norm:
energy = normalize_energy(energy)
erbs = np.flipud(calc_erbs(low_freq, fs, n_cochlear_filters))
avg_energy = np.mean(energy, axis=2)
total_energy = np.sum(avg_energy)
AC_energy = np.sum(avg_energy, axis=1)
AC_perc = AC_energy*100/total_energy
AC_perc_cumsum=np.cumsum(np.flipud(AC_perc))
K90perc_idx = np.where(AC_perc_cumsum>90)[0][0]
BW = erbs[K90perc_idx]
cutoffs = calc_cutoffs(mod_filter_cfs, fs, 2)[0]
if (BW > cutoffs[4]) and (BW < cutoffs[5]):
Kstar=5
elif (BW > cutoffs[5]) and (BW < cutoffs[6]):
Kstar=6
elif (BW > cutoffs[6]) and (BW < cutoffs[7]):
Kstar=7
elif (BW > cutoffs[7]):
Kstar=8
return np.sum(avg_energy[:, :4])/np.sum(avg_energy[:, 4:Kstar]), energy
def srmr(x, fs, n_cochlear_filters=23, low_freq=125, min_cf=4, max_cf=128, fast=True, norm=False):
wLengthS = .256
wIncS = .064
# Computing gammatone envelopes
if fast:
mfs = 400.0
gt_env = fft_gtgram(x, fs, 0.010, 0.0025, n_cochlear_filters, low_freq)
else:
cfs = centre_freqs(fs, n_cochlear_filters, low_freq)
fcoefs = make_erb_filters(fs, cfs)
gt_env = np.abs(hilbert(erb_filterbank(x, fcoefs)))
mfs = fs
wLength = np.ceil(wLengthS*mfs)
wInc = np.ceil(wIncS*mfs)
# Computing modulation filterbank with Q = 2 and 8 channels
mod_filter_cfs = compute_modulation_cfs(min_cf, max_cf, 8)
MF = modulation_filterbank(mod_filter_cfs, mfs, 2)
n_frames = 1 + (gt_env.shape[1] - wLength)//wInc
w = hamming(wLength+1)[:-1] # window is periodic, not symmetric
energy = np.zeros((n_cochlear_filters, 8, n_frames))
for i, ac_ch in enumerate(gt_env):
mod_out = modfilt(MF, ac_ch)
for j, mod_ch in enumerate(mod_out):
mod_out_frame = segment_axis(mod_ch, wLength, overlap=wLength-wInc, end='pad')
energy[i,j,:] = np.sum((w*mod_out_frame[:n_frames])**2, axis=1)
if norm:
energy = normalize_energy(energy)
erbs = np.flipud(calc_erbs(low_freq, fs, n_cochlear_filters))
avg_energy = np.mean(energy, axis=2)
total_energy = np.sum(avg_energy)
AC_energy = np.sum(avg_energy, axis=1)
AC_perc = AC_energy*100/total_energy
AC_perc_cumsum=np.cumsum(np.flipud(AC_perc))
K90perc_idx = np.where(AC_perc_cumsum>90)[0][0]
BW = erbs[K90perc_idx]
cutoffs = calc_cutoffs(mod_filter_cfs, fs, 2)[0]
if (BW > cutoffs[4]) and (BW < cutoffs[5]):
Kstar=5
elif (BW > cutoffs[5]) and (BW < cutoffs[6]):
Kstar=6
elif (BW > cutoffs[6]) and (BW < cutoffs[7]):
Kstar=7
elif (BW > cutoffs[7]):
Kstar=8
return np.sum(avg_energy[:, :4])/np.sum(avg_energy[:, 4:Kstar]), energy
def calculate_heinz2001_firing_rate(_input, fs, cfs, **kwargs):
""" Runs the Heinz et al. (2001) auditory nerve simulation and return firing rates
Implements the Heinz et al. (2001) auditory nerve model. This model contains the following steps:
- A gammatone frontend is implemented via the gammatone package (https://github.com/detly/gammatone)
- A saturating nonlinearity simulating the actions of the inner hair cells (IHC) is applied
- The IHC responses are lowpass filtered with 7 first-order Butterworth filters
- Auditory nerve responses to the IHC inputs are simulated - this stage is implemented via Numba for speed.
The implementation described in Heinz et al. (2001) is a slightly simplified version of three-stage
diffusion as in Westerman and Smith (1988).
Most of the parameter descriptions below in the inline documentation are taken directly from Heinz et al. (2001).
Args:
_input (ndarray): 1-dimensional ndarray containing an acoustic stimulus in pascals
fs (int): sampling rate in Hz
cfs (ndarray): ndarray containing characteristic frequencies at which to simulate responses
Returns:
output (ndarray): output array of instantaneous firing rates, of shape (n_cf, n_samp)
Warnings:
- Arguments passed via **kwargs are silently unused
References:
Heinz, M. G., Colburn, H. S., and Carney, L. H. (2001). "Evaluating auditory performance limits: I.
One-parameter discrimination using a computational model for the auditory nerve." Neural Computation, 13(10).
2273-2316.
Westerman, L. A., & Smith, R. L. (1988). A diffusion model of the transient response of the cochlear inner hair
cell synapse. The Journal of the Acoustical Society of America, 83(6), 2266-2276.
"""
# Calculate peripheral filter outputs
bm = filters.erb_filterbank(_input, filters.make_erb_filters(fs, cfs))
# Apply saturating nonlinearity
K = 1225 # controls sensitivity
beta = -1 # sets 3:1 asymmetric bias
ihc = (np.arctan(K * bm + beta) - np.arctan(beta)) / (np.pi / 2 -
np.arctan(beta))
# Apply lowpass filter
[b, a] = butter(1, 4800 / (fs / 2))
for ii in range(7):
ihc = lfilter(b, a, ihc, axis=1)
# Apply auditory nerve + neural adaptation stage
dims = ihc.shape
C_I = np.zeros_like(ihc) # immediate concentration ("spikes/volume")
C_L = np.zeros_like(ihc) # local concentration ("spikes/volume")
return _calculate_heinz2001_rate_internals(dims, fs, ihc, C_I, C_L)
def get_gfcc(self, signal, ccST=1, ccEND=23):
"""
Get GFCC feature.
"""
erb_filterbank = filters.erb_filterbank(numpy.array(signal),
self.erb_filter)
inData = erb_filterbank[10:, :]
[chnNum, frmNum] = numpy.array(inData).shape
mtx = self.dct_matrix(chnNum)
outData = numpy.matmul(mtx, inData)
outData = outData[ccST:ccEND, :]
gfcc_feat = numpy.array(
[numpy.mean(data_list) for data_list in outData]).copy()
return gfcc_feat
def _computeSingleFrameFeature(self,sig):
'''Feature computation for a single time-series frame/segment
Args:
sig (numpy array): The signal segment for which feature will be computed
Returns:
feature (numpy array): Computed feature vector
單個時間序列幀/段的特徵計算
(只限 ”SubEnv” 子帶包絡(Sub-band envelopes)特徵計算)
- 輸入變數 : sig (numpy array)
- 輸出變數 : feature (numpy array)
'''
if self.name=='SubEnv':
'''Sub-band envelopes feature computation 子帶包絡特徵計算'''
#Computing sub-band signals /計算子帶信號
timeRes=self.dimensions[0]
numBands=self.dimensions[1]
low_cut_off=2#lower cut off frequency = 2Hz /較低的截止頻率= 2Hz
centre_freqVals = centre_freqs(self.samplerate,numBands,low_cut_off)
fcoefs = make_erb_filters(self.samplerate, centre_freqVals, width=1.0)
y = erb_filterbank(sig, fcoefs)
subenv = np.array([]).reshape(timeRes,0)
for i in range(numBands):
subBandSig=y[i,:]
analytic_signal = hilbert(subBandSig)
amp_env = np.abs(analytic_signal)
np.nan_to_num(amp_env)
#amp_env=resampy.resample(amp_env, len(amp_env), timeRes, axis=-1)#resampy library used resampling /resampy庫使用重新取樣
#resampling may lead to unexpected computation errors, /重新採樣可能會導致意外的計算錯誤，
#I prefered average amplitudes for short-time windows /我更喜歡短時間窗口的平均幅度
downSampEnv=np.zeros((timeRes,1))
winSize=int(len(amp_env)/timeRes)
for ind in range(timeRes):
downSampEnv[ind]=np.log2(np.mean(amp_env[ind*winSize:(ind+1)*winSize]))
subenv=np.hstack([subenv,downSampEnv])
#removing mean and normalizing /刪除均值和正常化
subenv=subenv-np.mean(subenv)
subenv=subenv/(np.max(np.abs(subenv)))
feature=subenv
else:
print('Error: feature '+self.name+' is not recognized')
feature=[]
return feature
d = 0.89 #amount to remove detected signal from residual
time = 100 #time to start analyzing audio
delta_time = 100 # time between analysis frames
notes = np.zeros(6, dtype=float)
input_signal = wave.read('16kHz_acTuned.wav')
fs = input_signal[0]
T = 1 / fs
audio = np.asarray(input_signal[1])
while time < (len(audio) - fs * time / 1000):
trim_audio = clip_audio(audio, time)
center_freqs = gt.erb_space(lowf, highf, num_freqs)
filt_coefs = gt.make_erb_filters(fs, center_freqs)
channels = gt.erb_filterbank(trim_audio, filt_coefs)
process_chan = np.empty(
[len(channels), len(channels[0]) * 2], dtype=complex)
mag_chan = np.empty_like(process_chan, dtype=float)
#Cochlea simulation
for idx in range(len(channels)):
process_chan[idx, :] = compression(channels[idx, :])
process_chan[idx, :] = half_wave_rectification(process_chan[idx, :])
low_cutoff = center_freqs[idx] * 1.5
process_chan[idx, :] = butter_lfilter(process_chan[idx, :], low_cutoff,
fs)
mag_chan[idx, :] = np.absolute(np.fft.fft(process_chan[idx, :]))
#summation of DFT magnitudes across each channel
def gammatone_bank(
wav: NDVar,
f_min: float,
f_max: float,
n: int,
integration_window: float = 0.010,
tstep: float = None,
location: str = 'right',
pad: bool = True,
name: str = None,
) -> NDVar:
"""Gammatone filterbank response
Parameters
----------
wav : NDVar
Sound input.
f_min : scalar
Lower frequency cutoff.
f_max : scalar
Upper frequency cutoff.
n : int
Number of filter channels.
integration_window : scalar
Integration time window in seconds (default 10 ms).
tstep : scalar
Time step size in the output (default is same as ``wav``).
location : str
Location of the output relative to the input time axis:
- ``right``: gammatone sample at end of integration window (default)
- ``left``: gammatone sample at beginning of integration window
- ``center``: gammatone sample at center of integration window
Since gammatone filter response depends on ``integration_window``, the
filter response will be delayed relative to the analytic envlope. To
ignore this delay, use `location='left'`
pad : bool
Pad output to match time axis of input.
name : str
NDVar name (default is ``wav.name``).
Notes
-----
Requires the ``fmax`` branch of the gammatone library to be installed:
$ pip install https://github.com/christianbrodbeck/gammatone/archive/fmax.zip
"""
from gammatone.filters import centre_freqs, erb_filterbank
from gammatone.gtgram import make_erb_filters
wav_ = wav
if location == 'left':
if pad:
wav_ = _pad_func(wav, wav.time.tmin - integration_window)
elif location == 'right':
# tmin += window_time
if pad:
wav_ = _pad_func(wav, tstop=wav.time.tstop + integration_window)
elif location == 'center':
dt = integration_window / 2
# tmin += dt
if pad:
wav_ = _pad_func(wav, wav.time.tmin - dt, wav.time.tstop + dt)
else:
raise ValueError(f"mode={location!r}")
fs = 1 / wav.time.tstep
if tstep is None:
tstep = wav.time.tstep
wave = wav_.get_data('time')
# based on gammatone library, rewritten to reduce memory footprint
cfs = centre_freqs(fs, n, f_min, f_max)
integration_window_len = int(round(integration_window * fs))
output_n_samples = floor((len(wave) - integration_window_len) * wav.time.tstep / tstep)
output_step = tstep / wav.time.tstep
results = []
for i, cf in tqdm(enumerate(reversed(cfs)), "Gammatone spectrogram", total=len(cfs), unit='band'):
fcoefs = np.flipud(make_erb_filters(fs, cf))
xf = erb_filterbank(wave, fcoefs)
results.append(aggregate(xf[0], output_n_samples, output_step, integration_window_len))
result = np.sqrt(results)
# package output
freq_dim = Scalar('frequency', cfs[::-1], 'Hz')
time_dim = UTS(wav.time.tmin, tstep, output_n_samples)
if name is None:
name = wav.name
return NDVar(result, (freq_dim, time_dim), name)
"""
Renders the given ``duration`` of audio from the audio file at ``path``
using the gammatone spectrogram function ``function``.
"""
fig = matplotlib.pyplot.figure()
axes = fig.add_axes([0.1, 0.1, 0.8, 0.8])
erb_filterbank = filters.erb_filterbank(sig, )
axes.set_title("erb_filterbank"+ str(erb_filterbank.shape)+" " + os.path.basename(new_file_name_path))
axes.set_xlabel("Time (s)")
axes.set_ylabel("Frequency")
print("erb_filterbank.shape", erb_filterbank.shape)
matplotlib.pyplot.plot(erb_filterbank)
matplotlib.pyplot.show()
ipdb.set_trace()
in_sig_mt = np.zeros_like(t)
for p in range(2, num_h + 1):
if p == mistuned_h:
in_sig_mt += np.cos(2 * np.pi * f0 * p * t * mistuning)
else:
in_sig_mt += np.cos(2 * np.pi * f0 * p * t)
in_sig += np.cos(2 * np.pi * f0 * p * t)
in_sig /= np.max(in_sig)
in_sig_mt /= np.max(in_sig_mt)
################################################################################
# SAC Peak Picking
################################################################################
f_c = gtf.erb_space(20., 5000., num=num_channels)
erb_coefs = gtf.make_erb_filters(f_s, f_c)
filt_channels = gtf.erb_filterbank(in_sig, erb_coefs)
filt_channels_mt = gtf.erb_filterbank(in_sig_mt, erb_coefs)
ac_channels = np.zeros((num_channels, w_size))
summary_ac = np.zeros(w_size, dtype=float)
ac_channels_mt = np.zeros((num_channels, w_size))
summary_ac_mt = np.zeros(w_size, dtype=float)
for k in range(num_channels):
ac_channels[k, :] = dsp.correlate(filt_channels[k, -w_size:],
filt_channels[k, -w_size:],
mode='same')
ac_channels_mt[k, :] = dsp.correlate(filt_channels_mt[k, -w_size:],
filt_channels_mt[k, -w_size:],
mode='same')
summary_ac += np.clip(ac_channels[k, :], 1, None) # half-wave rectify
summary_ac_mt += np.clip(ac_channels_mt[k, :], 0,
def GetFilteredOutputFromArray(array, FILTERBANK_COEFFICIENTS):
# gammatone library needs a numpy array
# Application of the filterbank to a vector
filteredMatrix = filters.erb_filterbank(array, FILTERBANK_COEFFICIENTS)
# Matrix of wavFile.getnframes() X 128 real values
return filteredMatrix
def predict(self, clean, mixture, noise):
# Computing gammatone envelopes
if self.fast:
mfs = 400.0
gt_env = fft_gtgram(mixture, self.fs, 0.010, 0.0025,
self.n_cochlear_filters, self.low_freq)
else:
cfs = centre_freqs(self.fs, self.n_cochlear_filters, self.low_freq)
fcoefs = make_erb_filters(self.fs, cfs)
gt_env = np.abs(hilbert(erb_filterbank(mixture, fcoefs)))
mfs = self.fs
wLength = np.ceil(self.wLengthS*mfs)
wInc = np.ceil(self.wIncS*mfs)
# Computing modulation filterbank with Q = 2 and 8 channels
mod_filter_cfs = compute_modulation_cfs(self.min_cf, self.max_cf, 8)
MF = modulation_filterbank(mod_filter_cfs, mfs, 2)
n_frames = np.ceil((gt_env.shape[1])/wInc)
w = hamming(wLength)
energy = np.zeros((self.n_cochlear_filters, 8, n_frames))
for i, ac_ch in enumerate(gt_env):
mod_out = modfilt(MF, ac_ch)
for j, mod_ch in enumerate(mod_out):
mod_out_frame = segment_axis(mod_ch, wLength, overlap=wLength-wInc, end='delay')
energy[i,j,:] = np.sum((w*mod_out_frame)**2, axis=1)
if self.norm:
peak_energy = np.max(np.mean(energy, axis=0))
min_energy = peak_energy*0.001
energy[energy < min_energy] = min_energy
energy[energy > peak_energy] = peak_energy
erbs = np.flipud(self.calc_erbs(self.low_freq, self.fs,
self.n_cochlear_filters))
avg_energy = np.mean(energy, axis=2)
total_energy = np.sum(avg_energy)
AC_energy = np.sum(avg_energy, axis=1)
AC_perc = AC_energy*100/total_energy
AC_perc_cumsum=np.cumsum(np.flipud(AC_perc))
K90perc_idx = np.where(AC_perc_cumsum>90)[0][0]
BW = erbs[K90perc_idx]
cutoffs = self.calc_cutoffs(mod_filter_cfs, self.fs, 2)[0]
if (BW > cutoffs[4]) and (BW < cutoffs[5]):
Kstar=5
elif (BW > cutoffs[5]) and (BW < cutoffs[6]):
Kstar=6
elif (BW > cutoffs[6]) and (BW < cutoffs[7]):
Kstar=7
elif (BW > cutoffs[7]):
Kstar=8
out = {'p': {
'srmr': np.sum(avg_energy[:, :4]) / np.sum(avg_energy[:, 4:Kstar])},
'avg_energy': avg_energy
}
return out
def compute(filepath, file):
modelpath = 'C:/Users/user/Desktop/cnn/data/model/M_uocSeq1SubEnv32by16_nASyn2000len_1000hopt.h5'
#dir='C:/Users/Lab606B/Desktop/result/'#txt 儲存路徑
#wildcard="txt"
# fileLabels=['1']
timeDim = 32
freqDim = 16
frameSizeMs = 2000
hopSizeMs = 1000
signal, samplerate = sf.read(filepath + file)
lenSigSamp = len(signal)
lenSigMs = 1000 * lenSigSamp / samplerate
lenSigMs = lenSigMs
startsMs = list(np.arange(0, lenSigMs - frameSizeMs, hopSizeMs))
stopsMs = [x + frameSizeMs for x in startsMs]
#windowing using segmentation info and performing feature extraction /使用分段信息進行窗口化並執行特徵提取
starts = [int(round(x * samplerate / 1000)) for x in startsMs]
stops = [int(round(x * samplerate / 1000)) for x in stopsMs]
globalInd = 0
allFeatures = np.zeros((1, timeDim, freqDim))
# allLabels=[]
#(無用) fileSegmentMap={}#map containing filename versus indexes of segments/features within all samples in this set /包含文件名的映射與此集合中所有樣本中的段/要素的索引
for ind in range(len(starts)):
segment = signal[starts[ind]:stops[ind]]
#applying windowing function to the segment /將窗口函數應用於段
segment = segment * create_window(
stops[ind] - starts[ind], 'tukey', r=0.08)
if (np.max(segment) > 0): #normalization /正規化
segment = segment / np.max(segment)
#feature=Feature._computeSingleFrameFeature(segment)
'''Sub-band envelopes feature computation 子帶包絡特徵計算'''
#Computing sub-band signals /計算子帶信號
low_cut_off = 2 #lower cut off frequency = 2Hz /較低的截止頻率= 2Hz
centre_freqVals = centre_freqs(samplerate, freqDim, low_cut_off)
fcoefs = make_erb_filters(samplerate, centre_freqVals, width=1.0)
y = erb_filterbank(segment, fcoefs)
subenv = np.array([]).reshape(timeDim, 0)
for i in range(freqDim):
subBandSig = y[i, :]
analytic_signal = hilbert(subBandSig)
amp_env = np.abs(analytic_signal)
np.nan_to_num(amp_env)
#amp_env=resampy.resample(amp_env, len(amp_env), timeRes(timeDim), axis=-1)#resampy library used resampling /resampy庫使用重新取樣
#resampling may lead to unexpected computation errors, /重新採樣可能會導致意外的計算錯誤，
#I prefered average amplitudes for short-time windows /我更喜歡短時間窗口的平均幅度
downSampEnv = np.zeros((timeDim, 1))
winSize = int(len(amp_env) / timeDim)
for ind in range(timeDim):
downSampEnv[ind] = np.log2(
np.mean(amp_env[ind * winSize:(ind + 1) * winSize]))
subenv = np.hstack([subenv, downSampEnv])
#removing mean and normalizing /刪除均值和正常化
subenv = subenv - np.mean(subenv)
subenv = subenv / (np.max(np.abs(subenv)))
feature = subenv
#adding computed feature /添加計算特徵
if globalInd == 0: #if this is the first feature assign it directly /如果這是第一個功能直接分配它
allFeatures[0] = feature
else: #add one more element in the feature vector and then assign /在特徵向量中添加一個元素，然後分配
allFeatures = np.vstack(
[allFeatures, np.zeros((1, timeDim, freqDim))])
allFeatures[globalInd] = feature
#(無用) #adding segment to file-segment map /將段添加到文件段映射
#(無用) if file in fileSegmentMap:#if file already exists, append segment /如果文件已存在，則追加段
#(無用) val=fileSegmentMap[file]
#(無用) val.append(globalInd)
#(無用) fileSegmentMap[file]=val
#(無用) else:#file does not exist in map, add the first file-segment map /文件在地圖中不存在，添加第一個文件段映射
#(無用) fileSegmentMap[file]=[globalInd]
#(無用) allLabels.append(fileLabels)
globalInd += 1
#(無用) allFeatures=allFeatures.reshape(allFeatures.shape[0],timeRes,numBands,1)
#(無用) allLabels=np.array(allLabels,dtype = np.int)
#(無用) allLabels = to_categorical(allLabels)
allFeatures = np.reshape(allFeatures,
[len(allFeatures), timeDim, freqDim, 1])
#(無用) with open(filepath+'Test_Features.pkl', 'wb') as f:
#(無用) pickle.dump(allFeatures, f, 1)
#(無用) with open(filepath+'Test_Labels.pkl' , 'wb') as f:
#(無用) pickle.dump(allLabels, f, 1)
#(無用) with open(filepath+'Test_Map.pkl', 'wb') as f:
#(無用) pickle.dump(fileSegmentMap, f, 1)
model = keras.models.load_model(modelpath)
y_probs = model.predict(allFeatures,
batch_size=allFeatures.shape[0],
verbose=0)
#normal = -1 = 0 ; abnormal = 1
normal = 0
abnormal = 0
for i in range(len(y_probs)):
if (y_probs[i, 0] > y_probs[i, 1]):
normal = normal + 1
else:
abnormal = abnormal + 1
if (normal > abnormal):
result = 'normal'
resultRate = normal / len(y_probs) * 100
elif (normal < abnormal):
result = 'abnormal'
resultRate = abnormal / len(y_probs) * 100
else:
result = 'not sure'
resultRate = 50
#建立txt檔
text_file_predict = open(
'C:/Users/user/Desktop/cnn/DataSpaceFoeFTP/Predict_Result/nxp/' +
file.replace('.wav', '') + ".txt",
"w",
encoding='utf-8')
#text_file_predict.write('test result(predict)\n')
text_file_predict.write('檔案:' + str(file))
text_file_predict.write('\n')
text_file_predict.write('\n診斷結果 =\t' + str(result))
text_file_predict.write('\n概率為 =\t' + str(resultRate) + '%')
text_file_predict.write('\n------------------------------------------\n')
# ListFilesToTxt(dir,file,wildcard, 1)
text_file_predict.close()
print('診斷結果為 : ', result)
print('機率為 : ', resultRate, '%')
import numpy as np
import gammatone.filters as gtf
import matplotlib.pyplot as plt
# make signal
f_s = 44100
dt = 1 / f_s
dur = 0.1
t = np.arange(0, dur, dt)
num_h = 8
f0 = 220.0
in_sig = np.zeros_like(t)
for p in range(1, num_h + 1):
in_sig += np.cos(2 * np.pi * f0 * p * t)
# set up gammatone stuff
erb_freqs = gtf.erb_space(100.0, 4000.0, num=100)
print(erb_freqs)
erb_coefs = gtf.make_erb_filters(f_s, erb_freqs)
filted = gtf.erb_filterbank(in_sig, erb_coefs)
for k, channel in enumerate(filted):
plt.plot(t, channel + 100 - k, color='k')
plt.show()
fold = 0
for x in range(1, len(meta)):
if int(meta[x][5]) == fold + 1:
filename, foldno, classID = meta[x][0], int(meta[x][5]), int(
meta[x][6])
s, sr = librosa.load('UrbanSound8K/audio/fold' + str(foldno) + '/' +
filename,
sr=44100)
fcoefs = filters.make_erb_filters(sr,
filters.centre_freqs(sr, 128, 40),
odr=4)
g = filters.erb_filterbank(s, fcoefs)
c = cp.asarray(g)
c = cp.power(c, 2)
if len(c[0]) // 66536 > 0 and len(c[0]) % 65536 < 65536 / 2:
nspecs = len(c[0]) // 65536
else:
nspecs = (len(c[0]) // 65536) + 1
if len(c[0]) < 65536 + 512:
c = cp.pad(c, ((0, 0), (0, 66048 - len(c[0]))),
'constant',
constant_values=0)
# For each spectrogram,
for n in range(nspecs):
fig = plt.figure()
ax1 = fig.add_subplot(1, 2, 2)
ax1.set_xscale("log")
ax1.stem(CFs, pitches, basefmt=" ")
skip = 10
ax1.set_xticks([cf for cf in CFs[::skip]])
ax1.grid("on", axis='y')
ax1.set_xlabel("CF of Adapative Template", size=16)
ax1.set_ylabel("Phase-locked firing rate (Hz)", size=16)
ax1.get_xaxis().set_major_formatter(matplotlib.ticker.ScalarFormatter())
ax1.tick_params(axis='both', which='both', labelsize=12)
## place-rate profile, place-rate coding
f_c = gtf.erb_space(100., 1600., num=100)
erb_coefs = gtf.make_erb_filters(f_s, f_c)
filt_channels = gtf.erb_filterbank(in_sig, erb_coefs)
channel_power = np.zeros(len(filt_channels))
for k in range(len(channel_power)):
channel_power[k] = np.sqrt(np.dot(filt_channels[k], filt_channels[k]))
channel_power /= np.max(channel_power)
ax2 = fig.add_subplot(1, 2, 1)
ax2.set_xscale("log")
skip = 10
ax2.set_xticks([cf for cf in f_c[::skip]])
ax2.stem(f_c, channel_power, basefmt=" ")
ax2.grid("on", axis='y')
ax2.set_xlabel("CF of Auditory Channel", size=16)
ax2.set_ylabel("Normalized Power/Firing Rate", size=16)
ax2.get_xaxis().set_major_formatter(matplotlib.ticker.ScalarFormatter())
ax2.tick_params(axis='both', which='both', labelsize=12)
def perform_gammatone(audio_samples, filter_coeffs):
return gt_filters.erb_filterbank(audio_samples, filter_coeffs)
