search

giulbia / baby_cry_detection

Recognition of baby cry audio signal
254 stars 117 forks source link

hello, happen error: raise NoBackendError() #2

Open qqingdou opened 7 years ago

qqingdou commented 7 years ago

when i run the train_set.py, the error happen bellow: data, sample_rate = file_reader.read_audio_file() File "/usr/lib/python2.7/site-packages/baby_cry_detection-1.0-py2.7.egg/pc_methods/__init__.py", line 35, in read_audio_file audio_data, sr = librosa.load(self.file_name, sr=44100, mono=True, duration=5) File "/usr/lib/python2.7/site-packages/librosa-0.5.0-py2.7.egg/librosa/core/audio.py", line 107, in load with audioread.audio_open(os.path.realpath(path)) as input_file: File "/usr/lib/python2.7/site-packages/audioread/__init__.py", line 115, in audio_open raise NoBackendError() audioread.NoBackendError

giulbia commented 7 years ago

Hi,

If I rember correctly, I had the same error once at reading files. The problem was a hidden file (.DS_store, l'équivalent Mac OS X du Thumbs.db on Windows) in the folder where I was attempting to read audio files. Removing it should solve the problem.

qqingdou commented 7 years ago

Hi, I am not find this file named .DS_store on Linux System..

giulbia commented 7 years ago

With ls -la you can find out if there's anything that is not an audio file in the folder and remove it.

qqingdou commented 7 years ago

i used ls -la command to search file in the folder, there is nothing except ogg files list. it may be other error, let me try again and solve it, thanks

qqingdou commented 7 years ago

Hi, i have resolved the error "audioread.NoBackendError", but another error happen bellow: Traceback (most recent call last): File "train_set.py", line 83, in main() File "train_set.py", line 63, in main avg_features = feature_engineer.feature_engineer(audio_data=data) File "/usr/lib/python2.7/site-packages/baby_cry_detection-1.0-py2.7.egg/pc_methods/feature_engineer.py", line 112, in feature_engineer ), axis=0) ValueError: all the input array dimensions except for the concatenation axis must match exactly

giulbia commented 7 years ago

So what was the problem for the first error ?

Which commit have you downloaded? I've made modifications, I can't say what's wrong like that.

qqingdou commented 7 years ago

I have installed ffmpeg, So the first error "audioread.NoBackendError" disappeared.

When i run “python train_set.py" again, the error happened below: ValueError: all the input array dimensions except for the concatenation axis must match exactly

giulbia commented 7 years ago

I think is the same issue someone else is having. Which commit have you cloned ?