4

I tried using the official Apple example: SpeekToMe

I edited the example in the following way to get the confidence levels:

        if let result = result {
            for t in result.transcriptions
            {
                for s in t.segments
                {
                    print("POSSIBLE TRANSCRIPTION: \(s.substring) confidence: \(s.confidence)")
                }

            }
            self.textView.text = result.bestTranscription.formattedString
            isFinal = result.isFinal
        }

Problem is the confidence levels are always = 0.

I found similar questions but setting the defaultTaskHint to dictation (or anything else) didn't help.

Does anyone have any suggestions on how to get the proper confidence values?

Community
  • 1
  • 1
deloki
  • 1,729
  • 2
  • 18
  • 26
  • 3
    I find that as long as ```result.isFinal=false```, confidence is ```0.0``` but once it's ```true```, I get different values. You will need to call ```audioEngine.stop()``` to do that (as far as I know) – Kreutzer Apr 13 '17 at 03:51
  • Same problem here. However, I sometimes get non-zero `confidence` values when `isFinal` is still `false`. – Mischa Dec 27 '18 at 09:12

0 Answers0