Speech Recognition Training Data


Appen has many years of experience in collecting speech data according to client specifications. We have successfully executed on projects ranging from 20 speakers in one language, to 20,000 speakers in dozens of languages. Clients often choose to have Appen transcribe their data and add a matched pronunciation dictionary before delivery so that the result is ready for integration with a speech-recognition system. Appen also offers licensable resources that are useful for rapid startup or the expansion of coverage of a speech-recognition system. Clients benefit from our intimate regional and linguistic knowledge and our ability to get the job done.