zlzhang1124/acousticfeatureextraction

Acoustic feature extraction using Librosa library and openSMILE toolkit.使用Librosa音频处理库和openSMILE工具包，进行简单的声学特征提取

acousticfeatureextraction's Issues

hi，╥﹏╥...

你好，

我看了一篇论文，里面说它用
Acoustic: We utilize the toolkit OpenSMILE [13] to extract the
utterance-level features with the configuration of INTERSPEECH2010
[42]. The extracted feature vector is in 1,582 dimension.
不知道你知道open提取的是1582维度的吗？就是是固定的吗

best，
jun

请教

你好，想问一下，为什么提出来的特征参数，只有一帧的参数？

直接运行acoustic_feature.py报错：AttributeError: Unknown property c

代码没有改错，音频文件默认使用的audios/audio_raw.wav
Traceback (most recent call last):
File "D:/Projects/Audio/Emotion/SER/AcousticFeatureExtraction/acoustic_feature.py", line 1076, in
rhythm_f.plot()
File "D:/Projects/Audio/Emotion/SER/AcousticFeatureExtraction/acoustic_feature.py", line 579, in plot
wave_ax.set_ylabel("Normalized Amplitude", c="b")
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\axes_axes.py", line 267, in set_ylabel
return self.yaxis.set_label_text(ylabel, fontdict, **kwargs)
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\axis.py", line 1562, in set_label_text
self.label.update(kwargs)
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\text.py", line 185, in update
super(Text, self).update(kwargs)
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\artist.py", line 888, in update
for k, v in props.items()]
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\artist.py", line 888, in
for k, v in props.items()]
File "D:\InstallPath\Develop\Anaconda3\5.3.1\envs\Renovamen\lib\site-packages\matplotlib\artist.py", line 881, in _update_property
raise AttributeError('Unknown property %s' % k)
AttributeError: Unknown property c

Process finished with exit code 1

Subchunk1Size: 12smilePcm: bogus wave/riff header or file in wrong format

请问你使用的opensmile的版本？我使用opensmile2.3.0最新版本，使用我们场景的音频会报错，报错的解决方案网上查了，请问除了Matlab转换以外能在工程代码里面解决吗？报错如下：
smilePcm: Riff: 46464952
Format: 45564157
Subchunk1ID: 20746d66
Subchunk2ID: 61640000
AudioFormat: 1
Subchunk1Size: 12smilePcm: bogus wave/riff header or file in wrong format ('D:\Projects\Audio\Emotion\SER\AcousticFeatureExtraction\audios/EP03_seq04_sc132.wav')! (maybe you are trying to read a 32-bit wave file which is not yet supported (new header type...)?)

可以计算声品质指标吗?

请教：如何提取带有时序信息的音频特征呢？

您好，非常感谢您的工作。请教一下，对于一个.wav的音频文件，应该如何提取出（seq,dim）这样的特征呢，目前我提取的是（dim）,也就说一个音频特征只提取出一个vector，但我想提取出一个矩阵，也就是带帧信息/时序信息的特征，应该用librosa哪一个工具呢？