Seun you need to double the money to get in more serious-minded programmers to burst this thing open!

dhtml18:
Seun you need to double the money to get in more serious-minded programmers to burst this thing open!

CAN you shut UP for once
from the first page to now you have the highest comments but what have you done ?

Jskelly11:
CAN you shut UP for once
from the first page to now you have the highest comments but what have you done ?

Make E no be like say person no do anything...


here is my submission... not promising 90% accuracy though... but i think with the right training file (wav ...Mono 16bit 16000hz) you should get close to that..

Share Project File:
https://drive.google.com/open?id=0B5eDDamBqCKsZjFldnlCQzNqb28


Peace !!!

Yes! that is what we are talking about!! A winner at long last!!!!

@Jskelly11, please jog down to Seun's house to call him over here!

dhtml18:
Yes! that is what we are talking about!! A winner at long last!!!!

@Jskelly11, please jog down to Seun's house to call him over here!

Naaaaaaaaaaaa

Jskelly11:
CAN you shut UP for once
from the first page to now you have the highest comments but what have you done ?

If I cannot carry first for coding, make I carry first for talking now? I cannot lose at both ends!!

@seun.. this would have being a cinch had I known earlier.. plaese confirm, can one puggin to voice-recognition APIs for this task?

voltron:
@seun.. this would have being a cinch had I known earlier.. plaese confirm, can one puggin to voice-recognition APIs for this task?

This is actually not a voice recognition task. You can use open source libraries that can be compiled on Windows. (It's speaker recognition)

Seun:
This is actually not a voice recognition task. You can use open source libraries that can be compiled on Windows.

voltron:
@seun.. this would have being a cinch had I known earlier.. plaese confirm, can one puggin to voice-recognition APIs for this task?

It is not too late, oya cinch it with immediate effect! We don reach page 7 already sef!

dhtml18:


It is not too late, oya cinch it with immediate effect! We don reach page 7 already sef!

ok.. let me look at the requirements again

jacob05:
Make E no be like say person no do anything...


here is my submission... not promising 90% accuracy though... but i think with the right training file (wav ...Mono 16bit 16000hz) you should get close to that..

Share Project File:
https://drive.google.com/open?id=0B5eDDamBqCKsZjFldnlCQzNqb28


Peace !!!

CC: Seun Files shared with you

Space booked

@jacob05: I guess this is my punishment for forgetting to state submission requirements. Pls submit entries as yourusername.zip as before.

Let me rush over to his house to call him right away, brb

Seun:
@jacob05: I guess this is my punishment for forgetting to state submission requirements. Pls submit entries as yourusername.zip as before.

Nairaland has a upload restriction.... project file approx 4.1MB

jacob05:
Nairaland has a upload restriction.... project file approx 4.1MB

Never mind that; apologies. How does your code work?

Seun:
Never mind that; apologies. How does your code work?

It implements the MARF, The Modular Audio Recognition Framework, to do Training and Recognition http://marf.sourceforge.net/.

The MARF comes with a bundle of FEATURE EXtraction Methods to sample audio for "speaker properties" and Classification Methods ..

The program stores "speaker properties" analysed using LPC (Linear predictive coding ) and MAHALANOBIS_DISTANCE Classification Method in the DB when a audio file is inputted for training with a default id 0f 1.

The MARF recognize method enable input of a sample audio file which is analysed for "speaker properties" using the set FEATURE EXtraction Methods and Classification Method and then matched with "speakers properties" in it repository.

MARF would always return a match so the program calculates the probability of the matched id and if it is high enough (used >=0.75) then the speaker is recognized.

Note: It is also a fork of a user ucungsr
Don't know how to explain much but hope I'm clear enough.

bitbucket:

https://bitbucket.org/pyjac/pres-buhari-voice-recognition/src

This is a real professor at work - but different from the other ones that have no real codes to offer. Mehn, these are new terminologies, i no sabi anything walahi!

oga o

Seun:
As a sequel to the signature contest, here's a more challenging contest for programmers who are true computer scientists:

Write a program that will read a sound recording (mp3 format) containing at least 10 seconds of a Nigerian politician talking, and then print "Yes" to the console if the person speaking in the sound recording is President Buhari, and "No" if the person speaking is not President Buhari.

You can use any common programming language that runs on Windows 7/8/10. Use only open source libraries that I can compile on Windows.

The prize is 50,000 naira and I will award it to the first contestant whose code achieves a high level of accuracy (~90%).

Good luck!

mr seun please help,I can't read my mentions,it always come blank.I need an urgent solution please

Is this competition still on or a winner has been declared already

The competition is still on so to speak as no one has come close to winning yet. . . .you will be surprised due to all the talks

Ahhhh ...i been dey think say dhtml18 go done win this stuff since Page 2 fa! @Seun please next competition should be for Web Devs.

I for don try, no time to check time. I am not even sure I can win sef, I am not that good, you guys are just overhyping me.

J

barackosama:
Ahhhh ...i been dey think say dhtml18 go done win this stuff since Page 2 fa! @Seun please next competition should be for Web Devs.

If oga Seun stages a competition for web developers, he most likely will make it compulsory to be done with python and not php

What does here? No solution yet?

Laffing in spanish. . .we ain't got programmers for nairaland! programmers finito!!