search

interactiveaudiolab / emphases

Crowdsourced and Automatic Speech Prominence Estimation
https://maxrmorrison.com/sites/emphasis
MIT License
14 stars 2 forks source link

string does not have attribute shape #42

Closed YuCaIb closed 7 months ago

YuCaIb commented 7 months ago

`from glob import glob import emphases deu_audios = glob('test_data/resampled/deu_audios_resampled_wav/*.wav')

temp= 'Wenn du jemanden magst, dann verlasse deine Komfortzone und frage sie.' audio_file = deu_audios[0]

alignment,scores = emphases.from_text_and_audio(text = temp, audio=audio_file , sample_rate=16000)`

I'm getting the error : 'str' object has no attribute 'shape'

I think it is because some miss arg pasting, I tried to track -> emhases\core line 205 , pyfoal line 54 and p2fa line 34. I couldn't find anything about text is converted something else. Maybe it is about audio is not pasting from torch audio

YuCaIb commented 7 months ago

Okey I'm not getting this errors anymore I figured that, I had to past txt files path instead and needed to make some more installations. but I will be asking another question in another issue.