search

audeering / opensmile

The Munich Open-Source Large-Scale Multimedia Feature Extractor
https://audeering.github.io/opensmile/
Other
584 stars 75 forks source link

About emobase2010.conf #77

Open zhanglina94 opened 1 year ago

zhanglina94 commented 1 year ago

I'm trying to use OpenSmile as feature extractor(using emobase2010.conf) and do some classification with that features.

And I used 'SMILExtract -C /workspace/tts/opensmile/config/emobase/emobase2010.conf -I ' + path + '/audio/' + i + ' -O ' + '/workspace/dataset/mer1/opensmile-mer1/' + i[:-4] + '.csv') '

Output is (MSG) [2] SMILExtract: openSMILE starting! (MSG) [2] SMILExtract: config file is: /workspace/tts/opensmile/config/emobase/emobase2010.conf (MSG) [2] cComponentManager: successfully registered 102 component types. (WRN) [2] instance 'lld.reader': Inconsistency in input level parameters. growDyn on first (https://github.com/naxingyu/opensmile/issues/1) input = 0, growDyn on input # 5 (pitch) = 1. This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [2] instance 'lld.reader': Inconsistency in input level parameters. isRb on first (https://github.com/naxingyu/opensmile/issues/1) input = 1, isRb on input # 5 (pitch) = 0. This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [1] instance 'lld.reader': Input level buffer sizes (levelconf.nT) are inconsistent. Level 'intens' has size 2 which is smaller than the max. input size of all input levels (10). This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [1] instance 'lld.reader': Input level buffer sizes (levelconf.nT) are inconsistent. Level 'mfcc' has size 2 which is smaller than the max. input size of all input levels (10). This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [1] instance 'lld.reader': Input level buffer sizes (levelconf.nT) are inconsistent. Level 'mspec2log' has size 2 which is smaller than the max. input size of all input levels (10). This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [1] instance 'lld.reader': Input level buffer sizes (levelconf.nT) are inconsistent. Level 'lsp' has size 2 which is smaller than the max. input size of all input levels (10). This might cause the processing to hang unpredictably or cause incomplete processing. (WRN) [1] instance 'lld2.reader': Input level buffer sizes (levelconf.nT) are inconsistent. Level 'pitchF' has size 10 which is smaller than the max. input size of all input levels (100). This might cause the processing to hang unpredictably or cause incomplete processing. (ERR) [1] cArffSink: Error opening file '/workspace/dataset/mer1/opensmile-mer1/KETI_MULTIMODAL_0000001239_04.csv' for writing (component instance 'arffsink', type 'cArffSink') (ERR) [1] instance 'arffsink': No output was written! (Maybe the input was too short to extract features from, or the config is broken?) Input is 2 seconds per wav file How do I solve this problem? Thanks in advance~

maxschmitt commented 2 months ago

There are two issues here:

  1. Concerning the error, it seems that the folder where the output file is written to does not exist.
  2. The output format is ARFF, not CSV. emobase2010.conf does not support CSV, however, it could be adapted adding a CSV sink component.

Furthermore, this is related to opensmile, not opensmile-python.