Open laaptu opened 10 years ago
I think there's something in common. audio fingerprinting is a method used to find an exact match. even if the same singer sing the same song, it will be different for the fingerprint algorithm.
For your situation, fingerprinting maybe useful if the alarm is exact same or have a limited number.
but music is difference with alarm songs, which makes the results unpredictable.
my english is rather poor, wish you can catch up :)
PS:I'm not using echo print algorithm, you can read the article (An Industrial-Strength Audio Search Algorithm) write by wang which used by shazam
and there is a matlab version of its impelemention hope will help. ref: http://www.ee.columbia.edu/~dpwe/e4896/practicals.html#prac12
and here is a very useful project for you, might help you lot just check it out. ref: https://code.google.com/p/musicg/
and is your project opensouce? wonna to follow your progress cause i'm investigating fingerprinting too.
Thanks for the reply. The fire alarm will be same i.e. identical meaning I have to check for only one sound. So,I guess fingerprinting will work. And can I use your project for my purpose i.e. does it handle my problem? After checking with echoprint I will go for musicg as there are only two libraries that I can use so far in my Android. And once the project is successful ,I will surely opensource the fingerprinting portion because I think it is most needed in Android community.
Lastly, if I encounter any bumps,could I ask you for some peer guidance?
Thanks for the reply and guidance.
On Fri, May 16, 2014 at 3:01 PM, happyxgang notifications@github.comwrote:
and is your project opensouce? wonna to follow your progress cause i'm investigating fingerprinting too.
— Reply to this email directly or view it on GitHubhttps://github.com/happyxgang/EchoPrintAndroid/issues/1#issuecomment-43312202 .
didn't see your post until now...
what i intended to do is to implement the matlab script version algorithm in java. I don't know if it can fit your purpose, and my programming skill is rather poor and the code is still under a very unstable state.
if you need help, just post it here. i'm new to this too, so it's will not be a guidance, more like a discussion : )
hope everything goes well.
This is not an issue but a general question. I have to develop an app where I have to identify smoke alarm sound by listening from microphone. I am not an expert on Audio Fingerprinting but came across the article using Echo Print along with your Github page. I am looking upon the tutorial right now,but couldn't figure it out well. So, a humble question: could this library be used for my purpose i.e. to identify smoke alarm sound. Any advice will be greatly appreciated