2

I used the OpenEars for my app.just recognize "a" to "z" in the alphabet. But it had a bad recognition in recognize alphabet than word.

So, how can i use my sound model to improve the recognition of OpenEars.

And how can I use OpenEars to recognize some special sound.

for example. I give OpenEars a dog sound and I want it to give me back "dog"

brightintro
  • 1,016
  • 1
  • 11
  • 16
ryuikuya
  • 75
  • 2
  • 7
  • Unfortunately, this is a task that all speech recognition struggles with because the alphabet is mostly words with one or two phonemes and little context, and a whole lot of alphabet letters sound extremely similar to each other (c, d, e, g, p...). – Halle Nov 29 '12 at 13:20
  • I’m very sad to hear that. I try another way like this @"a",//always reconigze to k @"bee", @"sea", – ryuikuya Dec 01 '12 at 13:57
  • but still some word get a bad recognize,I hope someone can fix it @"a",//reconigze to "k" @"ee",//reconigze "t" or "bee" @"o o",//"r" @"r",//"i" @"v",//"ee" @"z",//"j"or"v" – ryuikuya Dec 01 '12 at 14:03
  • I don't understand what your approach is (this looks to me like it would cause more issues, not fewer), but if you want more assistance why not go to the OpenEars forums instead at http://www.politepix.com/forums/openears since your question is about the kind of subjective ongoing troubleshooting that isn't a great fit for Stack Overflow. – Halle Dec 01 '12 at 16:01
  • ok，I will discuss my quetion in politepix.com/forums/openears.thank you – ryuikuya Dec 02 '12 at 15:10

1 Answers1

2

So this is a two part question which might be better to the community split up. OpenEars from what I understand is best served as using words in the dictionary. If you want it to recognize alphabet letters I would try and use the phonetic spelling of each letter instead of using just the letter. So instead of using 'f' use "ef".

As for the second part of the question, you might be able to recognize specific types of dogs which go "ruff" but smaller dogs with more of a "yip!" would have to be added to the initial dictionary as well.

I would get the demo app and really just experiment with these words.

Evan Anger
  • 712
  • 1
  • 5
  • 24
  • Thank you for your answer.its very helpFull. Can you tell me how can I add the initial dictionary to openEar. – ryuikuya Nov 29 '12 at 06:00
  • OpenEars developer here. Just wanted to clarify that you can't recognize dog utterances using speech recognition with an English-language acoustic model by entering the English-language onomatopoeia for the sounds that dogs make. It might be possible to create an acoustic model with dog sound entries in the corpus that could indicate when a dog is barking, but that is more like a research project than a "how do I do it" SO question. – Halle Nov 29 '12 at 13:17
  • Also, and I'm sorry to be picky because I think it's a decent answer but I want to make sure this is a good resource for someone reading it later, the built-in language model generator will make the correct phonetic dictionary entries for the letters of the alphabet so it isn't necessary to first transcribe them. A more effective approach might be to emulate people who need to spell words under difficult environmental factors and use words to indicate letters (alpha, bravo, charlie, etc). – Halle Nov 29 '12 at 13:23
  • thank you for you answer.I think it's helpful. but, I maybe try another way to work out my problem. – ryuikuya Nov 29 '12 at 15:03
  • @Halle thanks for the clarification and I apologize the dog answer portion was more tongue-in-cheek. I couldn't help myself. I agree it is more of a research question. – Evan Anger Nov 29 '12 at 15:42
  • 1
    @EvanAnger np at all, thanks for being a good sport about my nitpicking. I definitely think the next OpenEars plugin needs to be the Bark-o-matic. – Halle Nov 29 '12 at 15:47