Tertioptus
asked on
Idea: Is it possible to auto-transcribe podcasts?
Hey Experts, I want to present and idea to you, and hopefully you can respond within following options:
1) Bad idea, go away loser
2) Good idea, already solved go here [add link]
3) Good idea, and unique, let's work on it
Okay, the podcast nation seems to be reaching it's peak, and my itunes container has become inundated with podcasts. I would like a to click on words in a "tags" portion of the "show notes" that would directly link me to positions in the mp3 file. The tags would be auto generated from a speech recognition engine, which would grab as many words is it could decipher (doesn't have to be all), convert them to text, and store them in a tags collection to compare against the show notes. Thus tags that appear in the show notes will be hyperlinks with an audio track position.
Why now? More than ever, people are downloading audio files because of the wide acceptance of the mp3 format, and now some are listening to mp3's for talk over music. You would rarely want skip portions of a song, but for "talk" this could be very beneficial. Imagine reading through you favorite podcasts' show notes, and you find that first topic bores you, but you see some about how "Zac Efron likes coloring when he's lonely". You could click on "lonely" blah blah blah... over explained, sorry i forgot who I was talking too.
It would be excellent if this is already out, yet I'm not opposed to investing my programming chops to solve the problem.
1) Bad idea, go away loser
2) Good idea, already solved go here [add link]
3) Good idea, and unique, let's work on it
Okay, the podcast nation seems to be reaching it's peak, and my itunes container has become inundated with podcasts. I would like a to click on words in a "tags" portion of the "show notes" that would directly link me to positions in the mp3 file. The tags would be auto generated from a speech recognition engine, which would grab as many words is it could decipher (doesn't have to be all), convert them to text, and store them in a tags collection to compare against the show notes. Thus tags that appear in the show notes will be hyperlinks with an audio track position.
Why now? More than ever, people are downloading audio files because of the wide acceptance of the mp3 format, and now some are listening to mp3's for talk over music. You would rarely want skip portions of a song, but for "talk" this could be very beneficial. Imagine reading through you favorite podcasts' show notes, and you find that first topic bores you, but you see some about how "Zac Efron likes coloring when he's lonely". You could click on "lonely" blah blah blah... over explained, sorry i forgot who I was talking too.
It would be excellent if this is already out, yet I'm not opposed to investing my programming chops to solve the problem.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER