File size: 2,432 Bytes
85b7206
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
import unittest

import models as mo
from constants import MODEL_NAME_DEFAULT
from tests.models.test_aimodels import helper_neural_asr


class TestModels(unittest.TestCase):

    def setUp(self):
        self.language_de = "de"
        self.language_en = "en"

    def test_getASRModel_de_whisper(self):
        from whisper_wrapper import WhisperASRModel
        asr = mo.getASRModel(self.language_de)
        self.assertIsInstance(asr, WhisperASRModel)
        asr_explicit = mo.getASRModel(self.language_de, model_name="whisper")
        self.assertIsInstance(asr_explicit, WhisperASRModel)

    def test_getASRModel_en_whisper(self):
        from whisper_wrapper import WhisperASRModel
        asr = mo.getASRModel(self.language_en)
        self.assertIsInstance(asr, WhisperASRModel)
        asr_explicit = mo.getASRModel(self.language_en, model_name="whisper")
        self.assertIsInstance(asr_explicit, WhisperASRModel)

    def test_neural_asr_de_whisper(self):
        self.maxDiff = None
        for model_name in [MODEL_NAME_DEFAULT, "whisper"]:
            audio_transcript, word_locations_in_samples = helper_neural_asr("de", model_name)
            assert audio_transcript == ' Hallo, wie geht es dir?'
            self.assertEqual(word_locations_in_samples, [
                {'end_ts': 5120.0, 'start_ts': 0.0, 'word': ' Hallo,'},
                {'end_ts': 10240.0, 'start_ts': 8640.0, 'word': ' wie'},
                {'end_ts': 13120.0, 'start_ts': 10240.0, 'word': ' geht'},
                {'end_ts': 16640.0, 'start_ts': 13120.0, 'word': ' es'},
                {'end_ts': 20160.0, 'start_ts': 16640.0, 'word': ' dir?'}
            ])

    def test_neural_asr_en_default(self):
        self.maxDiff = None
        for model_name in [MODEL_NAME_DEFAULT, "whisper"]:
            audio_transcript, word_locations_in_samples = helper_neural_asr("en", model_name)
            assert audio_transcript == ' Hi there, how are you?'
            self.assertEqual(word_locations_in_samples, [
                {'end_ts': 2240.0, 'start_ts': 0.0, 'word': ' Hi'},
                {'end_ts': 4800.0, 'start_ts': 2240.0, 'word': ' there,'},
                {'end_ts': 9280.0, 'start_ts': 7360.0, 'word': ' how'},
                {'end_ts': 11200.0, 'start_ts': 9280.0, 'word': ' are'},
                {'end_ts': 13760.0, 'start_ts': 11200.0, 'word': ' you?'}
            ])


if __name__ == '__main__':
    unittest.main()