From 5d5ef15ab701b6a190bd852b21ee9655ad73a226 Mon Sep 17 00:00:00 2001
From: Bernardo Henz <bernardohenz@gmail.com>
Date: Thu, 1 Aug 2019 21:00:55 -0300
Subject: [PATCH] -data-aug via additive and multiplicative noise in
 feature-space

---
 util/feeding.py | 10 +++++++++-
 util/flags.py   |  7 +++++++
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/util/feeding.py b/util/feeding.py
index ba11ebb0..fd8c400d 100644
--- a/util/feeding.py
+++ b/util/feeding.py
@@ -15,7 +15,7 @@ from tensorflow.python.ops import gen_audio_ops as contrib_audio
 from util.config import Config
 from util.logging import log_error
 from util.text import text_to_char_array
-
+from util.flags import FLAGS
 
 def read_csvs(csv_files):
     source_data = None
@@ -47,6 +47,14 @@ def audiofile_to_features(wav_filename):
     decoded = contrib_audio.decode_wav(samples, desired_channels=1)
     features, features_len = samples_to_mfccs(decoded.audio, decoded.sample_rate)
 
+
+    if FLAGS.data_aug_features_multiplicative > 0:
+        features = features*tf.random.normal(mean=1, stddev=FLAGS.data_aug_features_multiplicative, shape=tf.shape(features))
+
+    if FLAGS.data_aug_features_additive > 0:
+        features = features+tf.random.normal(mean=0.0, stddev=FLAGS.data_aug_features_additive, shape=tf.shape(features))
+
+
     return features, features_len
 
 
diff --git a/util/flags.py b/util/flags.py
index e1eb1788..3dc58fdd 100644
--- a/util/flags.py
+++ b/util/flags.py
@@ -21,6 +21,13 @@ def create_flags():
     f.DEFINE_integer('feature_win_step', 20, 'feature extraction window step length in milliseconds')
     f.DEFINE_integer('audio_sample_rate', 16000, 'sample rate value expected by model')
 
+    # Data Augmentation
+    # ================
+
+    f.DEFINE_float('data_aug_features_additive', 0, 'std of the Gaussian additive noise')
+    f.DEFINE_float('data_aug_features_multiplicative', 0, 'std of normal distribution around 1 for multiplicative noise')
+
+
     # Global Constants
     # ================