About
I am a deep learning researcher specializing in Text-To-Speech (TTS) technologies. At Cerence, I design and develop advanced TTS models with a focus on expressiveness, stability, and efficiency, contributing to both research innovation and real-world product impact. I hold a PhD in Electronics and Telecommunications Engineering from Politecnico di Torino, and my background includes extensive work in speech processing. I am passionate about advancing synthetic speech and voice technologies through principled, scalable approaches.
CV
- Current: Principal Research Scientist, Cerence (2023–Present)
- Past: Senior Research Engineer, Cerence | Senior AI Engineer, Unicredit | Postdoc, Politecnico di Torino
- Education: PhD in Electronics and Telecommunications Engineering, Politecnico di Torino (2016)
- Download Full CV (PDF)
Selected Publications
- 2025: Eta-WavLM: Efficient speaker identity removal in self-supervised speech representations using a simple linear equation (ACL)
- 2024: Enhancing Polyglot Voices by Leveraging Cross-Lingual Fine-Tuning in Any-to-One Voice Conversion (EMNLP)
- 2020: BioMetricNet: Deep unconstrained face verification through learning of metrics regularized onto Gaussian distributions (ECCV)
Contact
Email: moc.liamg@atsemt.ti
Website: www.matteotesta.it
LinkedIn: linkedin.com/in/matteo-testa-822a9811