search

piotrkawa / audio-deepfake-adversarial-attacks

Implementation of "Defense against Adversarial Attacks on Audio DeepFake Detection"
MIT License
50 stars 4 forks source link

pre-trained model #5

Open Tianming40 opened 8 months ago

Tianming40 commented 8 months ago

Dear author, I would like to ask you whether this project provides the pre-trained model you used in your paper. I want to use the model trained in this paper to evaluate a wild data set, but the training cost is a bit expensive for me.

best regards, Ming

piotrkawa commented 3 months ago

Hi, unfortunately we do not provide pre-trained models, however training procedure can easily allow to recreate the results. Alternatively, due to computational requirements, you could try using smaller batch size, however this will naturally affect the final performance.

Best regards, Piotr