audio-domain synthesis

dnn/lpcnet.py

@@ -4,6 +4,7 @@ import math
 from keras.models import Model
 from keras.layers import Input, LSTM, CuDNNGRU, Dense, Embedding, Reshape, Concatenate, Lambda, Conv1D, Multiply, Bidirectional, MaxPooling1D, Activation
 from keras import backend as K
+from keras.initializers import Initializer
 from mdense import MDense
 import numpy as np
 import h5py
@@ -14,6 +15,30 @@ pcm_bits = 8
 pcm_levels = 2**pcm_bits
 nb_used_features = 38
 
+class PCMInit(Initializer):
+    def __init__(self, gain=.1, seed=None):
+        self.gain = gain
+        self.seed = seed
+
+    def __call__(self, shape, dtype=None):
+        num_rows = 1
+        for dim in shape[:-1]:
+            num_rows *= dim
+        num_cols = shape[-1]
+        flat_shape = (num_rows, num_cols)
+        if self.seed is not None:
+            np.random.seed(self.seed)
+        a = np.random.uniform(-1.7321, 1.7321, flat_shape)
+        #a[:,0] = math.sqrt(12)*np.arange(-.5*num_rows+.5,.5*num_rows-.4)/num_rows
+        #a[:,1] = .5*a[:,0]*a[:,0]*a[:,0]
+        a = a + np.reshape(math.sqrt(12)*np.arange(-.5*num_rows+.5,.5*num_rows-.4)/num_rows, (num_rows, 1))
+        return self.gain * a
+
+    def get_config(self):
+        return {
+            'gain': self.gain,
+            'seed': self.seed
+        }
 
 def new_wavernn_model():
     pcm = Input(shape=(None, 1))
@@ -35,6 +60,10 @@ def new_wavernn_model():
         cpcm = pcm
         cpitch = pitch
 
+    embed = Embedding(256, 128, embeddings_initializer=PCMInit())
+    cpcm = Reshape((-1, 128))(embed(pcm))
+
+
     cfeat = fconv2(fconv1(feat))
 
     rep = Lambda(lambda x: K.repeat_elements(x, 160, 1))