Official implementation of the paper "Head-Related Transfer Function Interpolation with a Spherical CNN". [arXiv]
We appreciate the reviews and comments from ICASSP2024. Adjustments have been made based on the feedback to address some errors and improve our paper. An updated version will be uploaded in December.
Our paper has been submitted to ICASSP 2024, and we intend to release the complete code after the results are announced on December 13th. Stay tuned for updates!
Here's a brief overview of the repository's structure:
-
dataset.py: [Coming Soon] The
dataset.pyfile will provide the necessary data preprocessing and loading functions. -
model.py: This file contains the implementation of the learning model used in this project. It includes layers and the model architecture.
-
train.py: [Coming Soon] The
train.pyfile will contain the code for training the model using the HUTUBS dataset.
@article{chen2023head,
title={Head-Related Transfer Function Interpolation with a Spherical CNN},
author={Chen, Xingyu and Ma, Fei and Zhang, Yile and Bastine, Amy and Samarasinghe, Prasanga N},
journal={arXiv preprint arXiv:2309.08290},
year={2023}
}
Cohen, T. S., Geiger, M., Köhler, J., & Welling, M. (2018). Spherical cnns. arXiv preprint arXiv:1801.10130.
Esteves, C., Allen-Blanchette, C., Makadia, A., & Daniilidis, K. (2018). Learning so (3) equivariant representations with spherical cnns. In Proceedings of the European Conference on Computer Vision (ECCV) (pp. 52-68).
Brinkmann, F., Dinakaran, M., Pelzer, R., Wohlgemuth, J. J., Seipel, F., Voss, D., ... & Weinzierl, S. (2019). The HUTUBS head-related transfer function (HRTF) database.