Merge branch 'master' of /home/ilml/Public/Repos/speech_scoring

README.md

@@ -0,0 +1,18 @@
+### Setup
+`. env/bin/activate` to activate the virtualenv.
+
+### Data Generation
+* update `OUTPUT_NAME` in *speech_samplegen.py* to create the dataset folder with the name
+* `python speech_samplegen.py` generates variants of audio samples
+
+### Data Preprocessing
+* `python speech_data.py` creates the training-testing data from the generated samples.
+* run `fix_csv(OUTPUT_NAME)` to create the fixed index of the dataset generated
+* `generate_sppas_trans(OUTPUT_NAME)` creates the SPPAS transcription(wav+txt) data
+* `$ (SPPAS_DIR)/bin/annotation.py -l eng -e csv --ipus --tok --phon --align --align -w ./outputs/OUTPUT_NAME/` creates the phoneme alignment csv files for all variants.
+* `create_seg_phonpair_tfrecords(OUTPUT_NAME)` creates the tfrecords files
+ with the phoneme level pairs of right/wrong stresses
+
+### Training
+* `python speech_model.py` trains the model with the training data generated.
+* `train_siamese(OUTPUT_NAME)` trains the siamese model with the generated dataset.

speech_samplegen.py

@@ -216,6 +216,9 @@ def generate_audio_for_text_list(text_list):
     closer()
 
 def generate_audio_for_stories():
+    '''
+    Generates the audio sample variants for the list of words in the stories
+    '''
     # story_file = './inputs/all_stories_hs.json'
     story_file = './inputs/all_stories.json'
     stories_data = json.load(open(story_file))
@@ -225,6 +228,10 @@ def generate_audio_for_stories():
     generate_audio_for_text_list(text_list)
 
 def generate_test_audio_for_stories(sample_count=0):
+    '''
+    Picks a list of words from the wordlist that are not in story words
+    and generates the variants
+    '''
     story_file = './inputs/all_stories_hs.json'
     # story_file = './inputs/all_stories.json'
     stories_data = json.load(open(story_file))