Wheres the best technology in the world around speech to text conversion right now? I want to most efficiently capture my thoughts and ideas, to persist a thought stream. For 1. realtime personal use speaking to my phone 2. Providing a saved memo audio from my phone getting it back as text 3. Programmatically working with audio, the best APIs and physical sensor hardware. All suggestions most welcome. Im finding myself having a lot of wisdom and ideas that I want to capture the last few days and my natural way of doing that for me is on my iPhone to hold it up to my mouth and use the voice recorder and just stream of consciousness let the thought words come out when Im in a good energy a good body position and let the ideas spring forth. But this presents a problem once I go back how do I get the information quickly and easily, where it is and have it in text discoverable in indexed form so I can quickly get back to it and use it. Edit, wow Ive just switched to using the voice recording feature inside IOS notes and Im actually pretty impressed its updating the text back to me in real time as I speak and I can feel out the best speaking cadence to use, to with work it and speed up and slow down those words as some get caught incorrectly, l can then see this technology going back and editing and in realtime Im realizing do I need to slow down and change my speech pattern because its not quite getting my New Zealand/English accent and it needs a bit more American? Also this appears to only be present on the iPhone not on the iPad or Mac hmm I wish it there also. My question is what is the best technology in the world right now around capturing the human voice converting it as quickly as possible to what text it is, showing it back to me in front of me so that I can observe and then Ill adjust my speed, my cadence to make a better result and so the technology learns my mannerisms, my speech patterns my accidents and gets better and better providing and converting that text, Im noticing that Siri when I get some to maximum threshold of them talking it pauses and resets, breaks my flow and then I have to restart again. One specific question I have a real need for now is how do I provide an already saved voice memo to this service and have it run over a whole file 40 minute file and provide back to me all of the text as quickly as possible? When converted text at different words and sentences there was identified by the software a known low probability of correctness of conversion for that segment present, then provide that meta information back to me, I will seek to that position in the audio and do the conversion myself using my own ears and mouth or pulling from my own memory or recreating anew. Many thanks all for your collective knowledge, wisdom and co-creation.
Posted on: Thu, 16 Oct 2014 13:48:12 +0000
Trending Topics
Recently Viewed Topics
© 2015