![]() David Byron, editorial director of Speech Technology magazine suggests a technique called "parroting": listening to a recording in real-time and repeating its text back into the microphone for the software to transcribe. Where currently available transcription technology can't handle multiple voices or background chaos, reliable software like Nuance's Dragon NaturallySpeaking (also an outgrowth of Reddy's lab at Carnegie Mellon) has become quite capable at trained single voices. Those who only want to transcribe podcasts might have better luck. For the five phone interviews conducted for this article, recorded via Skype, only one subject spoke slowly and clearly enough to even register as recognizably transcribed text, with an error rate of roughly 15 percent. You can play recorded audio on your computer, and the system will do its best to make the proper text appear in a Google Doc. Google Made a Chatbot That Debates the Meaning of Life ArrowĪ freely available voice transcription tool is likewise built-in to Google Docs for those who would care to experiment. (Though with Mattel's voice-recognition-driven Hello Barbie that listens to the children playing with it, the dystopia might already be here.) Researchers say that functional transcription is only a matter of time, though the amount of time remains a very open question. It would usher a dystopia for others, providing a new form of textual panopticon. It would be a fantasy come true for researchers. When solved on a broad scale, it is a problem that might unlock vast archives of oral histories, make podcasts easier to consume for speed-readers (tl dl), and be a world-changing boon for journalists everywhere, liberating precious hours of sweet life. However, the task of providing accurate transcriptions of long blocks of actual human conversation remains beyond the abilities of even today's most advanced software. Our phones and smart home devices can understand fairly complex commands, thanks to self-teaching recurrent neural nets and other 21st century wonders. Sure, voice dictation for documents has been conquered by Nuance's Dragon software. In an age when technology companies routinely introduce new forms of everyday magic, one problem that remains seemingly unsolved is that of long-form transcription.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |