About me
I am a postdoctoral researcher at the Hearing Systems section at the Technical University of Denmark. I wrote my PhD thesis on deep learning-based speech enhancement between 2020 and 2023 at the Technical University of Denmark under the supervision of Tobias May and Tommy Sonne Alstrøm. My research interests include audio signal processing, deep learning-based speech enhancement and source separation, the generalization of learning-based systems, and model-based hearing loss compensation.
As a kid, I grew up in Colombia, Ivory Coast, Turkey, and Peru. I can speak English, French, Spanish, and some Danish. I play the bass in band, collect vinyl records, and occasionally DJ at parties. I like playing video games, especially Dota 2 in which I am a highly ranked player. Other hobbies include tennis, electronics, photography, and chess. I love dogs.
Publications
- P. Gonzalez, Z.-H. Tan, J. Østergaard, J. Jensen, T. S. Alstrøm, and T. May, “Investigating the Design Space of Diffusion Models for Speech Enhancement”, in IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2024.
- P. Gonzalez, Z.-H. Tan, J. Østergaard, J. Jensen, T. S. Alstrøm, and T. May, “The Effect of Training Dataset Size on Discriminative and Diffusion-Based Speech Enhancement Systems”, in IEEE Signal Processing Letters, 2024. [Audio examples]
- P. Gonzalez, Z.-H. Tan, J. Østergaard, J. Jensen, T. S. Alstrøm, and T. May, “Diffusion-Based Speech Enhancement in Matched and Mismatched Conditions Using a Heun-Based Sampler”, in IEEE International Conference on Acoustics, Speech and Signal Processing, 2024.
- P. Gonzalez, T. S. Alstrøm, and T. May, “Assessing the Generalization Gap of Learning-Based Speech Enhancement Systems in Noisy and Reverberant Environments”, in IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2023.
- P. Gonzalez, T. S. Alstrøm, and T. May, “On Batching Variable Size Inputs for Training End-to-End Speech Enhancement Systems”, in IEEE International Conference on Acoustics, Speech and Signal Processing, 2023.