speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

update repo

Browse Source
This commit is contained in:
嘉渊 2023-06-16 10:06:18 +08:00
parent 3bab35253d
commit 831a3f123c
1 changed files with 13 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
tests/test_sv_inference_pipeline.py
View File

@ -19,31 +19,21 @@ class TestXVectorInferencePipelines(unittest.TestCase):
task=Tasks.speaker_verification,
model='damo/speech_xvector_sv-zh-cn-cnceleb-16k-spk3465-pytorch'
)
# 提取不同句子的说话人嵌入码
rec_result = inference_sv_pipline(
audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_enroll.wav')
enroll = rec_result["spk_embedding"]
rec_result = inference_sv_pipline(
audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_same.wav')
same = rec_result["spk_embedding"]
# the same speaker
rec_result = inference_sv_pipline(audio_in=(
'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_enroll.wav',
'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_same.wav'))
assert abs(rec_result["scores"][0] - 0.85) < 0.1 and abs(rec_result["scores"][1] - 0.14) < 0.1
logger.info(f"Similarity {rec_result['scores']}")
rec_result = inference_sv_pipline(
audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_different.wav')
different = rec_result["spk_embedding"]
# different speaker
rec_result = inference_sv_pipline(audio_in=(
'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_enroll.wav',
'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/sv_example_different.wav'))
assert abs(rec_result["scores"][0] - 0.0) < 0.1 and abs(rec_result["scores"][1] - 1.0) < 0.1
logger.info(f"Similarity {rec_result['scores']}")
# 对相同的说话人计算余弦相似度
sv_threshold = 0.9465
same_cos = np.sum(enroll * same) / (np.linalg.norm(enroll) * np.linalg.norm(same))
same_cos = max(same_cos - sv_threshold, 0.0) / (1.0 - sv_threshold) * 100.0
logger.info("Similarity: {}".format(same_cos))
assert int(same_cos) == 85
# 对不同的说话人计算余弦相似度
diff_cos = np.sum(enroll * different) / (np.linalg.norm(enroll) * np.linalg.norm(different))
diff_cos = max(diff_cos - sv_threshold, 0.0) / (1.0 - sv_threshold) * 100.0
logger.info("Similarity: {}".format(diff_cos))
assert int(diff_cos) == 0.0
if __name__ == '__main__':
unittest.main()
unittest.main()