fix format && params

MAINTAINERS

@@ -1 +1,2 @@
 Hui Zhang <zhtclz@foxmail.com>
+Chengyun Deng <deng_chengyun@126.com>

delta/data/feat/speech_feature.py

@@ -15,11 +15,9 @@
 # ==============================================================================
 ''' speech feat entrypoint unittest'''
 import os
-
 import numpy as np
 import delta.compat as tf
 from absl import logging
-
 from delta.data.feat import speech_ops
 from delta.layers.ops import py_x_ops
 from delta.data.feat import python_speech_features as psf
@@ -86,7 +84,15 @@ def _freq_feat_graph(feat_name, **kwargs):
         spec = py_x_ops.spectrum(
             waveforms[:, 0],
             tf.cast(sample_rate, tf.dtypes.float32),
-            output_type=1)  #output_type: 1, power spec; 2 log power spec
+            window_length=0.025,
+            frame_length=0.010,
+            output_type=1,
+            snip_edges=1,
+            raw_energy=1,
+            preEph_coeff=0.97,
+            window_type='povey',
+            remove_dc_offset=True,
+            is_fbank=False)  #output_type: 1, power spec; 2 log power spec
         spec = tf.sqrt(spec)
         # shape must be [T, D, C]
         spec = tf.expand_dims(spec, -1)

delta/data/feat/speech_feature_test.py

@@ -16,12 +16,10 @@
 ''' speech feature entrypoint unittest'''
 import os
 from pathlib import Path
-
 import librosa
 import numpy as np
 import delta.compat as tf
 from absl import logging
-
 from delta.data.feat import speech_ops
 from delta.data.feat import speech_feature
 from delta import PACKAGE_ROOT_DIR

delta/data/frontend/analyfiltbank.py

@@ -15,12 +15,10 @@
 # ==============================================================================
 
 import delta.compat as tf
-
 from delta.layers.ops import py_x_ops
 from delta.utils.hparam import HParams
 from delta.data.frontend.base_frontend import BaseFrontend
 
-
 class Analyfiltbank(BaseFrontend):
 
   def __init__(self, config: dict):

delta/data/frontend/fbank.py

@@ -37,7 +37,7 @@ def params(cls, config=None):
         --frame_length				: Hop length in seconds. (float, default = 0.010)
         --snip_edges				  : If 1, the last frame (shorter than window_length) will be cutoff. If 2, 1 // 2 frame_length data will be padded to data. (int, default = 1)
         ---raw_energy				  : If 1, compute frame energy before preemphasis and windowing. If 2,  compute frame energy after preemphasis and windowing. (int, default = 1)
-        --preEph_coeff				: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
+        --preeph_coeff				: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
         --window_type				  : Type of window ("hamm"|"hann"|"povey"|"rect"|"blac"|"tria"). (string, default = "povey")
         --remove_dc_offset		: Subtract mean from waveform on each frame (bool, default = true)
         --is_fbank					  : If true, compute power spetrum without frame energy. If false, using the frame energy instead of the square of the constant component of the signal. (bool, default = true)
@@ -57,7 +57,7 @@ def params(cls, config=None):
     sample_rate = 16000
     snip_edges = 2
     raw_energy = 1
-    preEph_coeff = 0.97
+    preeph_coeff = 0.97
     window_type = 'povey'
     remove_dc_offset = True
     is_fbank = True
@@ -73,7 +73,7 @@ def params(cls, config=None):
     hparams.add_hparam('sample_rate', sample_rate)
     hparams.add_hparam('snip_edges', snip_edges)
     hparams.add_hparam('raw_energy', raw_energy)
-    hparams.add_hparam('preEph_coeff', preEph_coeff)
+    hparams.add_hparam('preeph_coeff', preeph_coeff)
     hparams.add_hparam('window_type', window_type)
     hparams.add_hparam('remove_dc_offset', remove_dc_offset)
     hparams.add_hparam('is_fbank', is_fbank)

delta/data/frontend/mfcc.py

@@ -36,7 +36,7 @@ def params(cls, config=None):
         --frame_length				: Hop length in seconds. (float, default = 0.010)
         --snip_edges				  : If 1, the last frame (shorter than window_length) will be cutoff. If 2, 1 // 2 frame_length data will be padded to data. (int, default = 1)
         ---raw_energy				  : If 1, compute frame energy before preemphasis and windowing. If 2,  compute frame energy after preemphasis and windowing. (int, default = 1)
-        --preEph_coeff				: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
+        --preeph_coeff				: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
         --window_type				  : Type of window ("hamm"|"hann"|"povey"|"rect"|"blac"|"tria"). (string, default = "povey")
         --remove_dc_offset		: Subtract mean from waveform on each frame (bool, default = true)
         --is_fbank					  : If true, compute power spetrum without frame energy. If false, using the frame energy instead of the square of the constant component of the signal. (bool, default = true)
@@ -58,7 +58,7 @@ def params(cls, config=None):
     sample_rate = 16000
     snip_edges = 2
     raw_energy = 1
-    preEph_coeff = 0.97
+    preeph_coeff = 0.97
     window_type = 'povey'
     remove_dc_offset = True
     is_fbank = True
@@ -75,7 +75,7 @@ def params(cls, config=None):
     hparams.add_hparam('sample_rate', sample_rate)
     hparams.add_hparam('snip_edges', snip_edges)
     hparams.add_hparam('raw_energy', raw_energy)
-    hparams.add_hparam('preEph_coeff', preEph_coeff)
+    hparams.add_hparam('preeph_coeff', preeph_coeff)
     hparams.add_hparam('window_type', window_type)
     hparams.add_hparam('remove_dc_offset', remove_dc_offset)
     hparams.add_hparam('is_fbank', is_fbank)

delta/data/frontend/plp_test.py

@@ -50,7 +50,7 @@ def test_plp(self):
            [0.052763, -0.271487, 0.011329, 0.025320, 0.012851]])
 
       self.assertEqual(tf.rank(plp_test).eval(), 2)
-      # Because the povey window (&& dithering=1.0) is used instead of the hamming window in spectrum.
+      # Because the povey window is used instead of the hamming window in spectrum.
       self.assertAllClose(plp_test.eval()[50:55, 5:10], output_true, rtol=1e-02, atol=1e-02)
 
 if __name__ == '__main__':

delta/data/frontend/spectrum.py

@@ -15,12 +15,10 @@
 # ==============================================================================
 
 import delta.compat as tf
-
 from delta.layers.ops import py_x_ops
 from delta.utils.hparam import HParams
 from delta.data.frontend.base_frontend import BaseFrontend
 
-
 class Spectrum(BaseFrontend):
 
   def __init__(self, config: dict):
@@ -36,7 +34,7 @@ def params(cls, config=None):
           --frame_length			: Hop length in seconds. (float, default = 0.010)
           --snip_edges			: If 1, the last frame (shorter than window_length) will be cutoff. If 2, 1 // 2 frame_length data will be padded to data. (int, default = 1)
           ---raw_energy			: If 1, compute frame energy before preemphasis and windowing. If 2,  compute frame energy after preemphasis and windowing. (int, default = 1)
-          --preEph_coeff			: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
+          --preeph_coeff			: Coefficient for use in frame-signal preemphasis. (float, default = 0.97)
           --window_type			: Type of window ("hamm"|"hann"|"povey"|"rect"|"blac"|"tria"). (string, default = "povey")
           --remove_dc_offset		: Subtract mean from waveform on each frame (bool, default = true)
           --is_fbank				: If true, compute power spetrum without frame energy. If false, using the frame energy instead of the square of the constant component of the signal. (bool, default = false)
@@ -50,7 +48,7 @@ def params(cls, config=None):
     sample_rate = 16000
     snip_edges = 2
     raw_energy = 1
-    preEph_coeff = 0.97
+    preeph_coeff = 0.97
     window_type = 'povey'
     remove_dc_offset = True
     is_fbank = False
@@ -62,7 +60,7 @@ def params(cls, config=None):
     hparams.add_hparam('sample_rate', sample_rate)
     hparams.add_hparam('snip_edges', snip_edges)
     hparams.add_hparam('raw_energy', raw_energy)
-    hparams.add_hparam('preEph_coeff', preEph_coeff)
+    hparams.add_hparam('preeph_coeff', preeph_coeff)
     hparams.add_hparam('window_type', window_type)
     hparams.add_hparam('remove_dc_offset', remove_dc_offset)
     hparams.add_hparam('is_fbank', is_fbank)
@@ -100,7 +98,7 @@ def call(self, audio_data, sample_rate=None):
             output_type=p.output_type,
             snip_edges=p.snip_edges,
             raw_energy=p.raw_energy,
-            preEph_coeff=p.preEph_coeff,
+            preEph_coeff=p.preeph_coeff,
             window_type=p.window_type,
             remove_dc_offset=p.remove_dc_offset,
             is_fbank=p.is_fbank)

delta/data/frontend/spectrum_test.py

@@ -18,12 +18,10 @@
 import numpy as np
 from pathlib import Path
 import delta.compat as tf
-
 from delta import PACKAGE_ROOT_DIR
 from delta.data.frontend.read_wav import ReadWav
 from delta.data.frontend.spectrum import Spectrum
 
-
 class SpectrumTest(tf.test.TestCase):
 
   def test_spectrum(self):

delta/layers/ops/kernels/add_rir_noise_aecres_1.2/BaseLib.cpp

@@ -14,78 +14,70 @@ See the License for the specific language governing permissions and
 limitations under the License.
 ==============================================================================*/
 
-#include <math.h>
-#include "typedefs_sh.h"
 #include "BaseLib.h"
+#include <math.h>
 #include <stddef.h>
+#include "typedefs_sh.h"
 
-void FFT(COMPLEX *pFFTData, int nFFTOrder)
-{
-        int n, i, nv2, j, k, le, l, le1, ip, nm1;
-        COMPLEX t, u, w;
-
-	    n = 1;
-	    for(i = 0; i < (int)nFFTOrder; i++)
-        {
-                n = n * 2;
-        }
-
-	    nv2 = n / 2;
-	    nm1 = n - 1;
-	    j = 1;
-
-	    for (i = 1; i <= nm1; i ++)
-	    {
-		        if (i < j)
-		        {
-			            t.real = pFFTData[i - 1].real;
-			            t.image = pFFTData[i - 1].image;
-			            pFFTData[i - 1].real = pFFTData[j - 1].real;
-			            pFFTData[i - 1].image = pFFTData[j - 1].image;
-			            pFFTData[j - 1].real = t.real;
-			            pFFTData[j - 1].image = t.image;
-		        }
-
-		        k = nv2;
-
-		        while (k < j)
-		        {
-			            j -= k;
-			            k /= 2;
-		        }
-		        j += k;
-	    }
-
-	    le = 1;
-
-	    for (l= 1; l <= (int)nFFTOrder; l++)
-	    {
-		        le *= 2;
-		        le1 = le / 2;
-		        u.real = 1.0f;
-		        u.image = 0.0f;
-		        w.real = (float) cos(PI / le1);
-		        w.image =(float) -sin(PI / le1);
-
-		        for (j = 1; j <= le1; j++)
-		        {
-			            for (i = j; i <= n; i += le)
-			            {
-				                ip = i + le1;
-				                t.real = pFFTData[ip - 1].real * u.real - pFFTData[ip - 1].image * u.image;
-				                t.image = pFFTData[ip - 1].real * u.image + pFFTData[ip - 1].image * u.real;
-				                pFFTData[ip - 1].real = pFFTData[i - 1].real - t.real;
-				                pFFTData[ip - 1].image = pFFTData[i - 1].image - t.image;
-				                pFFTData[i - 1].real = t.real + pFFTData[i - 1].real;
-				                pFFTData[i - 1].image = t.image + pFFTData[i - 1].image;
-			            }
-
-			            t.real = u.real * w.real - u.image * w.image;
-			            t.image = u.image * w.real + u.real * w.image;
-			            u.real = t.real;
-			            u.image = t.image;
-		        }
-        }
+void FFT(COMPLEX *pFFTData, int nFFTOrder) {
+  int n, i, nv2, j, k, le, l, le1, ip, nm1;
+  COMPLEX t, u, w;
+
+  n = 1;
+  for (i = 0; i < (int)nFFTOrder; i++) {
+    n = n * 2;
+  }
+
+  nv2 = n / 2;
+  nm1 = n - 1;
+  j = 1;
+
+  for (i = 1; i <= nm1; i++) {
+    if (i < j) {
+      t.real = pFFTData[i - 1].real;
+      t.image = pFFTData[i - 1].image;
+      pFFTData[i - 1].real = pFFTData[j - 1].real;
+      pFFTData[i - 1].image = pFFTData[j - 1].image;
+      pFFTData[j - 1].real = t.real;
+      pFFTData[j - 1].image = t.image;
+    }
+
+    k = nv2;
+
+    while (k < j) {
+      j -= k;
+      k /= 2;
+    }
+    j += k;
+  }
+
+  le = 1;
+
+  for (l = 1; l <= (int)nFFTOrder; l++) {
+    le *= 2;
+    le1 = le / 2;
+    u.real = 1.0f;
+    u.image = 0.0f;
+    w.real = (float)cos(PI / le1);
+    w.image = (float)-sin(PI / le1);
+
+    for (j = 1; j <= le1; j++) {
+      for (i = j; i <= n; i += le) {
+        ip = i + le1;
+        t.real =
+            pFFTData[ip - 1].real * u.real - pFFTData[ip - 1].image * u.image;
+        t.image =
+            pFFTData[ip - 1].real * u.image + pFFTData[ip - 1].image * u.real;
+        pFFTData[ip - 1].real = pFFTData[i - 1].real - t.real;
+        pFFTData[ip - 1].image = pFFTData[i - 1].image - t.image;
+        pFFTData[i - 1].real = t.real + pFFTData[i - 1].real;
+        pFFTData[i - 1].image = t.image + pFFTData[i - 1].image;
+      }
+
+      t.real = u.real * w.real - u.image * w.image;
+      t.image = u.image * w.real + u.real * w.image;
+      u.real = t.real;
+      u.image = t.image;
+    }
+  }
 }
-
-

delta/layers/ops/kernels/add_rir_noise_aecres_1.2/BaseLib.h

@@ -14,10 +14,10 @@ See the License for the specific language governing permissions and
 limitations under the License.
 ==============================================================================*/
 
-#ifndef  __BASELIB_H_
-#define  __BASELIB_H_
+#ifndef __BASELIB_H_
+#define __BASELIB_H_
 
-#include	"typedefs_sh.h"
+#include "typedefs_sh.h"
 
 void FFT(COMPLEX *pFFTData, int nFFTOrder);
 

delta/layers/ops/kernels/add_rir_noise_aecres_1.2/CAdd_All.cpp

@@ -14,25 +14,17 @@ See the License for the specific language governing permissions and
 limitations under the License.
 ==============================================================================*/
 
-#include "conv.h"
 #include "CAdd_All.h"
-
-CAdd_All::CAdd_All()
-{
-    st_rir = conv_init(16000, 0); 
-}
-
-CAdd_All::~CAdd_All()
-{
-    conv_exit(st_rir);
-}
+#include "conv.h"
 
-int CAdd_All::add_rir(void* st, short* inputdata, int inputdata_length, 
-                short* outputdata, int* outputdata_size, char* filelist)
-{
-    int ret;
-    ret = conv_process(st, inputdata, inputdata_length, 
-            outputdata, outputdata_size, filelist);
-    return ret;
-}
+CAdd_All::CAdd_All() { st_rir = conv_init(16000, 0); }
+
+CAdd_All::~CAdd_All() { conv_exit(st_rir); }
 
+int CAdd_All::add_rir(void* st, short* inputdata, int inputdata_length,
+                      short* outputdata, int* outputdata_size, char* filelist) {
+  int ret;
+  ret = conv_process(st, inputdata, inputdata_length, outputdata,
+                     outputdata_size, filelist);
+  return ret;
+}