Pooja199 / SpeechClassification

This model is used to distinguish between genuine and fake audio samples. Automated Speaker Verification systems are prone to replay attacks. The objective is to provide countermeasures to replay attacks in ASV systems.
1 stars 4 forks source link

Example #1

Open ovninosa opened 7 years ago

ovninosa commented 7 years ago

Hi,

Can you make an example to work with this usings wavs?

Thanks Jave

Pooja199 commented 7 years ago

Can you be more specific, so i can explain it better.

Thanks Pooja

On Sun, Oct 8, 2017 at 12:23 PM, javenosa notifications@github.com wrote:

Hi,

Can you make an example to work with this usings wavs?

Thanks Jave

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Pooja199/SpeechClassification/issues/1, or mute the thread https://github.com/notifications/unsubscribe-auth/AMS5_Kbrb-_IGakjKPMN2zWuFw3vZPbpks5sqHFQgaJpZM4PxnTF .

ovninosa commented 7 years ago

Yes, I want to test your code with a voices dataset audio files (wav) to analyze the genuinity. The evaluation result is a score or boolean?

Can you make an example with that? or at least a guide? I see a few paths on your code but are related to your pc.

Thanks Jave

ovninosa commented 6 years ago

@Pooja199 can you help me with this?

Thanks!

Pooja199 commented 6 years ago

The result is a score. You first perform cqcc though which you get the scores for your training data set. Use this data (path) for your check.

Regards, Pooja

On Sat, Nov 25, 2017 at 7:28 AM, javenosa notifications@github.com wrote:

@Pooja199 https://github.com/pooja199 can you help me with this?

Thanks!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Pooja199/SpeechClassification/issues/1#issuecomment-346912672, or mute the thread https://github.com/notifications/unsubscribe-auth/AMS5_HDpFsb3oizyvdvYupbzZ6rFKd3bks5s53RIgaJpZM4PxnTF .

ovninosa commented 6 years ago

@Pooja199 I saw this paths over your code.

With the files of this repo is enough to test it? Or I need another file from your personal pc?

Thanks

        % set paths to the wave files and protocols
        pathToDatabase = fullfile('C:\Users\Pooja\Desktop\baseline_CM','ASVspoof2017_train_dev');
        trainProtocolFile = fullfile('C:\Users\Pooja\Desktop\baseline_CM\ASVspoof2017_train_dev', 'protocol', 'ASVspoof2017_train.trn');
        devProtocolFile = fullfile('C:\Users\Pooja\Desktop\baseline_CM\ASVspoof2017_train_dev', 'protocol', 'ASVspoof2017_dev.trl');

        pathToDatabase = fullfile('C:\Users\Shrirag\Desktop','ASV Trimmed Data');

        evalProtocolFile = fullfile('C:\Users\Shrirag\Desktop','ASV Trimmed Data','ASVspoof2017_eval.trl');
Pooja199 commented 6 years ago

Hey. This is enough

On Wed, Mar 28, 2018, 2:43 AM javenosa notifications@github.com wrote:

@Pooja199 https://github.com/Pooja199 I saw this paths over your code.

With the files of this repo is enough to test it? Or I need another file from your personal pc.

Thanks

  % set paths to the wave files and protocols
  pathToDatabase = fullfile('C:\Users\Pooja\Desktop\baseline_CM','ASVspoof2017_train_dev');
  trainProtocolFile = fullfile('C:\Users\Pooja\Desktop\baseline_CM\ASVspoof2017_train_dev', 'protocol', 'ASVspoof2017_train.trn');
  devProtocolFile = fullfile('C:\Users\Pooja\Desktop\baseline_CM\ASVspoof2017_train_dev', 'protocol', 'ASVspoof2017_dev.trl');

  pathToDatabase = fullfile('C:\Users\Shrirag\Desktop','ASV Trimmed Data');

  evalProtocolFile = fullfile('C:\Users\Shrirag\Desktop','ASV Trimmed Data','ASVspoof2017_eval.trl');

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Pooja199/SpeechClassification/issues/1#issuecomment-376677334, or mute the thread https://github.com/notifications/unsubscribe-auth/AMS5_OC-GWTaIiNkFS54sdY74Fnc9RS9ks5tiqtjgaJpZM4PxnTF .