FreeDV
/
phasenn
mirror of https://github.com/drowe67/phasenn.git
1
0
Fork 0
Code Issues Releases Wiki Activity

fixed custom loss function - good matches and even overfitting - but still probs with UV

master
David 2019-12-09 20:22:05 +10:30
parent 5c19633779
commit c1484a76c1
2 changed files with 31 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
phasenn_train.py
View File

@ -56,8 +56,10 @@ for i in range(nb_samples):
amp[i,bin] = np.log10(A[i,m])
#phase_rect[i,2*bin] = np.max((1,amp[i,bin]))*np.cos(phase[i,m])
#phase_rect[i,2*bin+1] = np.max((1,amp[i,bin]))*np.sin(phase[i,m])
phase_rect[i,2*bin] = amp[i,bin]*np.cos(phase[i,m])
phase_rect[i,2*bin+1] = amp[i,bin]*np.sin(phase[i,m])
#phase_rect[i,2*bin] = amp[i,bin]*np.cos(phase[i,m])
#phase_rect[i,2*bin+1] = amp[i,bin]*np.sin(phase[i,m])
phase_rect[i,2*bin] = np.cos(phase[i,m])
phase_rect[i,2*bin+1] = np.sin(phase[i,m])
# our model
model = models.Sequential()
@ -66,24 +68,28 @@ model.add(layers.Dense(4*pairs, activation='relu'))
model.add(layers.Dense(pairs))
model.summary()
# custom loss function
# custom loss function - we only care about (cos,sin) outputs at the
# non-zero positions in the sparse y_true vector. To avoid driving the
# other samples to 0 we use a sparse loss function. The normalisation
# term accounts for the time varying number of no-zero samples.
def sparse_loss(y_true, y_pred):
mask = K.cast( K.not_equal(y_pred, 0), dtype='float32')
mask = K.cast( K.not_equal(y_true, 0), dtype='float32')
n = K.sum(mask)
return K.sum(K.square((y_pred - y_true)*mask))/n
# testing custom loss function
x = Input(shape=(None,))
y = Input(shape=(None,))
loss_func = K.Function([x, y], [sparse_loss(x, y)])
assert loss_func([[[1,1,1]], [[0,2,0]]]) == np.array([1])
y_true = Input(shape=(None,))
y_pred = Input(shape=(None,))
loss_func = K.Function([y_true, y_pred], [sparse_loss(y_true, y_pred)])
assert loss_func([[[0,1,0]], [[2,2,2]]]) == np.array([1])
assert loss_func([[[1,1,0]], [[3,2,2]]]) == np.array([2.5])
assert loss_func([[[0,1,0]], [[0,2,0]]]) == np.array([1])
# fit the model
from keras import optimizers
sgd = optimizers.SGD(lr=0.8, decay=1e-6, momentum=0.9, nesterov=True)
sgd = optimizers.SGD(lr=0.05, decay=1e-6, momentum=0.9, nesterov=True)
model.compile(loss=sparse_loss, optimizer=sgd)
history = model.fit(amp, phase_rect, batch_size=nb_batch, epochs=args.epochs)
history = model.fit(amp, phase_rect, batch_size=nb_batch, epochs=args.epochs, validation_split=0.1)
model.save(args.nnout)
# measure error in angle over all samples
@ -113,11 +119,14 @@ def sample_time(r, phase):
s = s + A[r,m]*np.cos(m*Wo[r]*range(-N,N) + phase[r,m])
return s
nb_plotsy = np.floor(np.sqrt(nb_plots)); nb_plotsx=nb_plots/nb_plotsy;
frames = np.array(args.frames,dtype=int)
nb_plots = frames.size
nb_plotsy = np.floor(np.sqrt(nb_plots)); nb_plotsx=nb_plots/nb_plotsy;
plt.figure(1)
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.legend(['train', 'valid'], loc='upper right')
plt.title('model loss')
plt.xlabel('epoch')
plt.show(block=False)

15
train.sh
View File

@ -4,9 +4,16 @@
PATH=$PATH:~/codec2/build_linux/src:~/codec2/build_linux/misc
speech=~/Downloads/train_8k.sw
if [ "$#" -eq 0 ]; then
echo "usage: ./train.sh rawFile [secs]"
fi
speech=$1
x=$(basename $speech)
base="${x%.*}"
sox -t .sw -r 8000 -c 1 $speech -t .sw - trim 0 6000 | c2sim - --modelout - | est_n0 -r > $base'_nolinear.model'
./phasenn_train.py $base'_nolinear.model' --frames 1572,1908,6792,9600,24536,25116 --epochs 200
if [ "$#" -eq 2 ]; then
sox -t .sw -r 8000 -c 1 $speech -t .sw - trim 0 $2 | c2sim - --modelout - | est_n0 -r > $base'_nolinear.model'
else
c2sim $speech --modelout - | est_n0 -r > $base'_nolinear.model'
fi
./phasenn_train.py $base'_nolinear.model' --frames 1572,1908,6792,9600,24536,25116 --epochs 10