Predicting binaural colouration using VGGish embeddings

McKenzie T, Wright A, Turner D, Lladó P. Predicting binaural colouration using VGGish embeddings. In: 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio. Audio Engineering Society; 2025. Paper 18. Available from: https://aes.org/publications/elibrary-page/?id=23007

@inproceedings{McKenzie2025_23007,
  author = {McKenzie, Thomas and Wright, Alec and Turner, Daniel and Lladó, Pedro},
  title = {{Predicting binaural colouration using VGGish embeddings}},
  booktitle = {2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio},
  note = {Paper 18},
  year = {2025},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23007}
}

TY  - CPAPER
TI  - Predicting binaural colouration using VGGish embeddings
AU  - McKenzie, Thomas
AU  - Wright, Alec
AU  - Turner, Daniel
AU  - Lladó, Pedro
T2  - 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio
M1  - Paper 18
PY  - 2025
DA  - 2025/09/02
UR  - https://aes.org/publications/elibrary-page/?id=23007
PB  - Audio Engineering Society
LA  - en
AB  - An initial feasibility study is presented exploring the use of a pre-trained feature extractor designed for large-scale audio classification applied to the task of predicting colouration between binaural signals. A multilayer perceptron (MLP) is trained to predict binaural colouration using feature embeddings obtained from the VGGish network and data from five previously conducted listening tests. The evaluation compares seven versions of the network, each trained using different data augmentation methods, along with three existing signal processing methods for predicting binaural colouration: basic spectral difference (BSD), log. spectral distance (LSD) and an auditory model for predicting binaural colouration (PBC-2). Results show that while the MLP networks are comparable to BSD and LSD, specific features relevant for colouration may be needed to compete against the more complex PBC-2.
ER  -