Theo Lepage
Ph.D. Student in A.I. → Self-Supervised Learning for Speaker Recognition
Paris, France
theolepage.com
theo@theolepage.com
Theo LepageEducation
Sorbonne Université (Ph.D. in Artificial Intelligence)
Paris, France
Nov. 2022 - Nov. 2025- Thesis: "Self-Supervised Learning for Speaker Recognition"
- Supervised by Dr. Réda Dehak and Pr. Thierry Géraud (LRE-EPITA)
- Supported by the French ANR 'APATE' project (Forensic Deepfakes Detection Toolbox)
École Pour l'Informatique et les Techniques Avancées - EPITA (M.Eng. in Computer Science)
Paris, FranceSep. 2017 - Sep. 2022- Major: IMAGE (Machine Learning, Deep Learning, Signal Processing, Probability & Statistics, Linear Algebra, Optimization)
- Specialization: Research (RDI) on Speaker Recognition (supervised by Dr. Réda Dehak) → publication at Interspeech 2022
Experience
Ph.D. Student at EPITA Research Laboratory (LRE)
Paris, FranceNov. 2022 - Nov. 2025- Published 8 peer-reviewed papers at top conferences and journals (Interspeech, Odyssey, IEEE TASLP, Speech Comm.)
- Developed and maintained sslsv, a toolkit for training and evaluating self-supervised speaker verification models
- Supervised an M.Eng. student leading to a co-authored publication at Interspeech 2024
- Teaching assistant at EPITA for "Intro to Deep Neural Networks" and "Python for Data Science" (2023–2025)
Research Scientist Intern at Siemens Healthineers
Princeton, USAFeb. 2022 - Sep. 2022- Focused on state-of-the-art deep learning models for MR images enhancement (denoising and super-resolution)
- Designed a CNN architecture that leverages the attention mechanism of Vision Transformers and recovers more details compared to the solution being used in the product
Software Developer Intern at CNRS
Paris, FranceSep. 2020 - Jan. 2021- Contributed to a real-time digital holography software (C++ / CUDA) used for retinal blood flow analysis in a medical setting
- Our work resulted in a 20x (500 to 10,000 FPS) speedup which improved substantially output images contrast and quality
- Our refactoring and the addition of unit tests improved the stability and allowed the project to become open source
- Founding member of the association 'Digital Holography' created to sustain the development of the software
Publications
Speech Communication, vol. 176, pp. 103333, 2026
Theo Lepage and Reda Dehak
Interspeech 2025, Aug. 2025, pp. 1098--1102
Theo Lepage and Reda Dehak
IEEE Transactions on Audio, Speech and Language Processing, vol. 33, pp. 2932--2945, 2025
Theo Lepage and Reda Dehak
The Automatic Speaker Verification Spoofing Countermeasures Workshop (ASVspoof 2024), Aug. 2024, pp. 72--78
Theophile Stourbe, Victor Miara, Theo Lepage, and Reda Dehak
Towards Supervised Performance on Speaker Verification with SSL by Leveraging Large-Scale ASR ModelsInterspeech 2024, Sept. 2024, pp. 2660--2664
Victor Miara, Theo Lepage, and Reda Dehak
The Speaker and Language Recognition Workshop (Odyssey 2024), Jun. 2024, pp. 38--42
Theo Lepage and Reda Dehak
Interspeech 2023, Aug. 2023, pp. 4708--4712
Theo Lepage and Reda Dehak
Interspeech 2022, Sept. 2022, pp. 4018--4022
Theo Lepage and Reda Dehak
Projects
Toolkit for training and evaluating Self-Supervised Learning (SSL) frameworks for Speaker Verification (SV).
SOTA method for self-supervised speaker verification leveraging a large-scale pretrained ASR model.
Talks
SSPS: Self-Supervised Positive Sampling for Robust Self-Supervised Speaker VerificationInterspeech 2025 — Rotterdam, The NetherlandsAug. 2025Self-Supervised Learning for Speaker RecognitionInria — Paris, FranceMar. 2025Self-Supervised Learning for Speaker RecognitionCallyope — Paris, FranceDec. 2024Exploring WavLM Back-ends for Speech Spoofing and Deepfake DetectionThe Automatic Speaker Verification Spoofing Countermeasures Workshop (ASVspoof 2024) — Kos, GreeceAug. 2024Additive Margin in Contrastive Self-Supervised Frameworks to Learn Discriminative Speaker RepresentationsThe Speaker and Language Recognition Workshop (Odyssey 2024) — Quebec, CanadaJun. 2024Experimenting with Additive Margins for Contrastive Self-Supervised Speaker VerificationInterspeech 2023 — Dublin, IrelandAug. 2023Label-Efficient Self-Supervised Speaker Verification With Information Maximization and Contrastive LearningInterspeech 2022 — Incheon, South KoreaSep. 2022Lightning Talks on SSL applied to Speaker and Language RecognitionLRDE — Paris, FranceJan. - Sep. 2021Teaching
Intro to Deep Neural NetworksEPITASpring 2023 - 2025Rust ProgrammingEPITASpring 2020Python for Data ScienceEPITASpring 2023 - 2025Unix / C ProgrammingEPITAFall 2019Skills and interests
Programming
C
C++
C#
Java
Python
PHP
JavaScript
Bash
Data Science
PyTorch
TensorFlow
Scikit-learn
NumPy
Pandas
Languages
English (TOEIC 905)
French (native)
Passions and interests
Science and AI
Robotics
Sailing & windsurfingAn up-to-date version of this document is available at theolepage.com/resume.