[Coeur]

Actually, https://raw.githubusercontent.com/cmusphinx/cmudict/master/cmudict.dict is even newer, but I haven’t tried it yet. And it has additional digits for each phonem: don’t know if it’s good or bad.

• This reply was modified 6 years, 5 months ago by Coeur.

[Coeur]

Thank you.

[Coeur]

Note that for compatibility with OpenEars, I had to replace the first space of each line with a tab:

Either with sed:

sed 's/ /'$'\t''/' cmudict-en-us.dict > LanguageModelGeneratorLookupList.text

Or with perl:

perl -pe 's/ /\t/' cmudict-en-us.dict > LanguageModelGeneratorLookupList.text

• This reply was modified 6 years, 5 months ago by Coeur.
• This reply was modified 6 years, 5 months ago by Coeur.

[Coeur]

I understand there is a concern about support. I’ll try to workaround that by keeping the same name but different folders:

/1/AcousticModelEnglish.bundle
/2/AcousticModelEnglish.bundle
/3/AcousticModelEnglish.bundle
etc.

[Coeur]

@Daniele, on OpenEars 2.041, I am successfully using OpenEars outside the mainBundle:

self.accousticModel = [[NSBundle bundleForClass:self.class] pathForResource:@”MyAcousticModel” ofType:@”bundle”];
[sphinxController runRecognitionOnWavFileAtPath:self.wavFilePath usingLanguageModelAtPath:self.languageModelPath dictionaryAtPath:self.dictionaryPath acousticModelAtPath:self.accousticModel languageModelIsJSGF:YES];

But I had trouble with version 2.502, as there are more restrictions than before at using custom models.

[Coeur]

heartbeatTimer issue occurs if you set your deployment target to iOS 5.1.1.

Workaround: set deployment target to at least iOS 6.0.

[Coeur]

Oh, thank you very much Halle.

> Starting OpenEars logging for OpenEars version 1.7 on 64-bit device: iPhone running iOS version: 8.400000

I got previously bullshited by someone telling me it was 2.x, ah ah. Hopefully I have now upgraded it to 2.041, thanks to your https://www.politepix.com/upgradeguide/

[Author]

Posts