My name is Theo Lepage. I hold a Ph.D. in Artificial Intelligence from Sorbonne University and an M.Eng. in Computer Science from EPITA. My research focuses on Deep Learning for Speech Processing, notably self-supervised learning, speaker and language recognition, and speech anti-spoofing & deepfake detection. Through my work, I aim to advance speech technologies for improving human-computer communication.
Alternatively, you can explore my profiles on:
Publications
Self-Supervised Learning for Speaker Recognition: A study and review
2026
SSPS: Self-Supervised Positive Sampling for Robust Self-Supervised Speaker Verification
2025
Self-Supervised Frameworks for Speaker Verification via Bootstrapped Positive Sampling
2025
IEEE Transactions on Audio, Speech and Language Processing
Theo Lepage and Reda Dehak
Towards Supervised Performance on Speaker Verification with Self-Supervised Learning by Leveraging Large-Scale ASR Models
2024
Additive Margin in Contrastive Self-Supervised Frameworks to Learn Discriminative Speaker Representations
2024
The Speaker and Language Recognition Workshop (Odyssey 2024)
Theo Lepage and Reda Dehak
See all publications
Projects
sslsv
35
9
Toolkit for training and evaluating Self-Supervised Learning (SSL) frameworks for Speaker Verification (SV).
wavlm_ssl_sv
8
0
SOTA method for self-supervised speaker verification leveraging a large-scale pretrained ASR model.
tensorflood
3
0
Tiny automatic differentiation (autodiff) engine for NumPy tensors implemented in Python.
Talks
SSPS: Self-Supervised Positive Sampling for Robust Self-Supervised Speaker Verification
Aug. 2025
Interspeech 2025 — Rotterdam, The Netherlands
Self-Supervised Learning for Speaker Recognition
Mar. 2025
Inria — Paris, France
See all talks
Teaching
Python for Data Science
Spring 2023 - 2025 @ EPITA
Intro to Deep Neural Networks
Spring 2023 - 2025 @ EPITA
Rust Programming
Spring 2020 @ EPITA
Unix / C Programming
Fall 2019 @ EPITA
© 2026 Theo Lepage