How to use Speech Recognition inside the iOS SDK? [closed]
Asked Answered
C

2

45

I know that there is no public API for the SIRI-Services, but is there an API for simple Speech-Recognition? So if I have a textfield and the user taps onto that textfield, a keyboard with the typically microphone button appears and if he pressed it the speech get recognized and transformed into a string object? Or is this button maybe presented by default?

Chapbook answered 29/5, 2013 at 11:14 Comment(9)
This below is two speech api politepix.com/openears && ispeech.orgTussock
Check Text to speech on iPhone and #12840171Proboscidean
This question is not a duplicate of TTS questions. This question is asking how to convert speech into text, not the other way around.Schwab
Checkout GIRI: github.com/jhurt/GIRI. Disclaimer: I wrote it.Sasin
You should tell us what languages shall be recognized.Drin
Apple introduced Speech Recognition in iOS10 -> let recognizer = SFSpeechRecognizer() let request = SFSpeechURLRecognitionRequest(url: audioFileURL) recognizer?.recognitionTask(with: request, resultHandler: { (result, error) in print (result?.bestTranscription.formattedString) })Northway
@TiagoAlmeida, Yeah, but code you put here that is for preRecorded audio filesUnni
@TiagoAlmeida: SFSpeechAudioBufferRecognitionRequest will be helpful for you. Its objects records and buffers the audio data while you are giving input. So your buffered data can be given to the SFSpeechURLRecognitionRequest.Bound
@TiagoAlmeida :mindbowser.com/blog/siri-speech-to-text refer this for more informationBound
W
43

There are many libraries availble. You can use any of them.

  1. openears // This is the best library

  2. VocalKit (Deprecated for open ears)

  3. TTS
  4. ispeech (Not free)

Hope it helps you.

NOTE:

if you download openears (which contains a sample project called "OpenEarsSampleApp") @efimovD mentions this

Check the code in view controller and you will see an array with possible commands. This thing detects commands! Not some talk. It listens and tries to compare what you've said with the words from array

Warrin answered 29/5, 2013 at 11:23 Comment(8)
It seems openear's sample app always thinks I am repeating the word "GO". Fails completely. Using iPhone 6+ with 8.1.2Kasi
You can also consider CeedVocal (not free) which has been used in many apps on iOS. github.com/creaceed/CeedVocal and creaceed.com/ceedvocalVivisection
Is TTS for speech recognition or only for text-to-speech? If only for the latter one, please remove it from the list.Drin
which one is best for indian accent ?Oney
I checked the docs and made an edit. TTS does indeed do TTS only and not recognition.Oden
@Kasi Yes, this is because you don' t understand what it does. Check the code in view controller and you will see an array with possible commands. This thing detects commands! Not some talk. It listens and tries to compare what you've said with the words from array. The best comparison will be displayed.Blether
ispeech.org/#/home claims iSpeech is free and open-source for Mobile use..?Sciential
You can find sample app for TTS over here ---> github.com/jonbarlo/iPhoneTTSSampleAppMarna
S
1

You can use the MindMeld iOS SDK provided by Expect Labs. It supports continuous (streaming) speech recognition on iOS. It also supports natural language understanding and content discovery for any custom content domain.

Try it out at: https://expectlabs.com/docs/sdks/ios/gettingStarted

Sulphuric answered 19/4, 2015 at 0:27 Comment(1)
To the suggester: you shouldn't do that, the link should be left as is, because the reader could understand where are they going to click.Shrivel

© 2022 - 2024 — McMap. All rights reserved.