Re:An Introduction to Sine-Wave Speech (2007)
by hellofunk on Thursday August 20, 2015 @02:52PM
(#10093844)
Crazy. As I listened to each example of the sin waves repeatedly, they went from gibberish to perfect sensical English. I think I sensed my neural nets learning the proper outcome from the same inputs. I think I am a machine.
Re:An Introduction to Sine-Wave Speech (2007)
by wanderfowl on Thursday August 20, 2015 @03:17PM
(#10094012)
Sine wave speech is an example I use in my early Linguistics classes to reinforce the position that speech perception is basically magic. Absolutely incredible that it works.
Re:An Introduction to Sine-Wave Speech (2007)
by xecax on Thursday August 20, 2015 @03:51PM
(#10094204)
It's a fantastic example, isn't it? I had to give a lightning presentation about a topic in linguistics for a class and the SWS synthesis demonstration grabbed the crowd's attention like nothing else.
Re:An Introduction to Sine-Wave Speech (2007)
by wanderfowl on Thursday August 20, 2015 @04:50PM
(#10094561)
Yeah, it's a real crowd pleaser, particularly if you play the individual components of the signal one-at-a-time first, so they hear nothing linguistic, and then combined, language!
It so nicely captures one of the key takeaways from my own line of research, which focuses on our robustness, as listeners, to problems and variability in the signal.
Re:An Introduction to Sine-Wave Speech (2007)
by janzer on Thursday August 20, 2015 @11:43PM
(#10095871)
Yes, I skipped the text at first and wanting to see what this sine wave speech sounded like just clicked on the first sample. I understood all but a word or two on the first play through. After reading the text I seem to do about the same on the other samples as well.
Interestingly for me, although standard hearing tests show I have good hearing (somewhat above average for my age despite experiencing minor tinnitus) I seem to be worse than normal at understanding speech in noisy environments (e.g. near loud engines).
Re:An Introduction to Sine-Wave Speech (2007)
by mlady on Thursday August 20, 2015 @09:33PM
(#10095525)
This seems like a principle that we've probably all experienced before with songs.
Just think of the last song you didn't know the words to and had to look up the lyrics. Next time you heard it, you heard the words clearer because you knew what you were looking for!
I experience this all the time and for years have just mumbled along to some parts to my favorite songs until I actually cared enough to see what they were saying.
It so nicely captures one of the key takeaways from my own line of research, which focuses on our robustness, as listeners, to problems and variability in the signal.
Interestingly for me, although standard hearing tests show I have good hearing (somewhat above average for my age despite experiencing minor tinnitus) I seem to be worse than normal at understanding speech in noisy environments (e.g. near loud engines).
Just think of the last song you didn't know the words to and had to look up the lyrics. Next time you heard it, you heard the words clearer because you knew what you were looking for!
I experience this all the time and for years have just mumbled along to some parts to my favorite songs until I actually cared enough to see what they were saying.