Skip to content

madhavlab/2023_interactiveMelEx

Repository files navigation

Interactive singing melody extraction based on active adaptation

The paper associated with these codes:

  • Kavya Ranjan Saxena and Vipul Arora. "Interactive singing melody extraction based on active adaptation." IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2024.

Trained models are in the folder weights

Step-by-step training:

STEP 1: Run the pretrain_basemodel.py file to pre-train the melody estimation model. The base model is present in the models.py file.
STEP 2: Run the pretrain_confmodel.py file to pre-train the confidence model. The confidence model is present in the models.py file.
STEP 3: Once we have obtained the pre-train base and confidence model, we apply active-meta-learning by running the active_meta_training.pyfile.
STEP 4: Once we have obtained the active-meta-trained model, we use it to adapt to the audios in the target domain by running the active_meta_testing.py file

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages