Merge branch 'master' of https://code.whiteblossom.net/malar/speech-scoring

speech_samplegen.py

@@ -14,7 +14,7 @@ import progressbar
 from generate_similar import similar_phoneme_phrase,similar_phrase
 from speech_tools import format_filename
 
-OUTPUT_NAME = 'story_all'
+OUTPUT_NAME = 'story_words_test'
 dest_dir = os.path.abspath('.') + '/outputs/' + OUTPUT_NAME + '/'
 dest_file = './outputs/' + OUTPUT_NAME + '.csv'
 
@@ -238,10 +238,25 @@ def generate_audio_for_stories():
     # story_file = './inputs/all_stories_hs.json'
     story_file = './inputs/all_stories.json'
     stories_data = json.load(open(story_file))
-    # word_list = [t[0] for i in stories_data.values() for t in i]
+    # text_list = [t[0] for i in stories_data.values() for t in i]
     text_list = [i.replace('-','') for g in stories_data.values() for i in g]
     generate_audio_for_text_list(text_list)
 
+def generate_test_audio_for_stories():
+    story_file = './inputs/all_stories_hs.json'
+    # story_file = './inputs/all_stories.json'
+    stories_data = json.load(open(story_file))
+    text_list = [t[0] for i in stories_data.values() for t in i]
+    # text_list = [i.replace('-','') for g in stories_data.values() for i in g]
+    word_list = [i.strip('\n_') for i in open('./inputs/wordlist.txt','r').readlines()]
+    text_set = set(text_list)
+    new_word_list = [i for i in word_list if i not in text_set and len(i) > 4]
+    # len(new_word_list)
+    test_words = new_word_list[:int(len(text_list)/5+1)]
+    generate_audio_for_text_list(test_words)
+
+
 if __name__ == '__main__':
+    generate_test_audio_for_stories()
     # generate_audio_for_text_list(['I want to go home','education'])
-    generate_audio_for_stories()
+    # generate_audio_for_stories()