Hi, I got Sphinx4 to work and recognize English :)
However, the accuracy is not so good yet. I am using StreamSpeechRecognizer and this:
configuration.setAcousticModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us"); configuration.setDictionaryPath("resource:/edu/cmu/sphinx/models/en-us/cmudict-en-us.dict"); configuration.setLanguageModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us.lm.bin");
Can I use better language models? Like those monsters here? https://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English/
I have 4 GB of RAM and want the best possible accuracy.
Please advise :-)
It is better to use kaldi DNN model instead, it is more accurate.
Log in to post a comment.
Hi, I got Sphinx4 to work and recognize English :)
However, the accuracy is not so good yet. I am using StreamSpeechRecognizer and this:
configuration.setAcousticModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us");
configuration.setDictionaryPath("resource:/edu/cmu/sphinx/models/en-us/cmudict-en-us.dict");
configuration.setLanguageModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us.lm.bin");
Can I use better language models? Like those monsters here? https://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English/
I have 4 GB of RAM and want the best possible accuracy.
Please advise :-)
It is better to use kaldi DNN model instead, it is more accurate.