PeopleofSuperintelligence

Researcher

Matthieu Perrot

Matthieu Perrot is a machine-learning scientist and a coauthor of the 2011 paper that presented scikit-learn, the Python machine-learning library. Trained at EPITA and ENS Cachan, he was a neuroimaging PhD student at NeuroSpin from 2006 to 2009 and published a method for recognizing cortical sulci and normalizing brain images in 2011. After research and engineering roles at DOSIsoft, the ICM Brain & Spine Institute and Philips, he served as head of AI at L'Oréal from 2016 to 2024. At L'Oréal, he coauthored GANs for face aging and makeup transfer and inverse-graphics methods for real-time virtual try-on. He is the founder and president of Evanova, an AI and computer-vision consulting company he started in December 2024.

Now at
Evanova
Works
48
Citations
70,565
h-index
16
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2000–2009

    From EPITA to a NeuroSpin PhD

    He studies advanced computer science and artificial intelligence at EPITA from 2000 to 2004, then takes a master's in computer vision, machine learning and applied mathematics at ENS Cachan from 2004 to 2005. From 2006 to 2009, he is a neuroimaging PhD student at ENS Cachan and a research engineer at CEA Saclay, working in NeuroSpin's computer-assisted neuroimaging laboratory (LNAO). His doctoral research focuses on automatic recognition of cortical sulci and spatial normalization.

  2. 2009–2014

    DOSIsoft and the CATI project

    He works as an R&D engineer at DOSIsoft from 2009 to 2011 on machine learning and medical imaging, then manages R&D for the CATI project at the ICM Brain & Spine Institute from 2011 to 2014. In July 2014, he coauthors a NeuroSpin paper on TV-Elastic Net penalties for logistic regression on structural MRI.

  3. 2011–2012

    Cortical sulci recognition and normalization

    In 2011, he publishes "Cortical sulci recognition and spatial normalization" in Medical Image Analysis with D. Rivière and J.-F. Mangin. In 2012, he coauthors a NeuroImage study on the effect of handedness on the shape of the central sulcus.

  4. 2011

    Coauthoring the scikit-learn paper

    In 2011, he coauthors "Scikit-learn: Machine Learning in Python" in the Journal of Machine Learning Research with Fabian Pedregosa, Gaël Varoquaux, Édouard Duchesnay and others. The paper presents scikit-learn, a Python module of machine-learning algorithms for medium-scale supervised and unsupervised problems, aimed at non-specialists.

  5. 2014–2017

    Research scientist at Philips

    After short data-science roles at Dailymotion and Twenga in 2014, he becomes a research scientist at Philips in Suresnes, working on AI, machine learning and deep learning R&D until 2016. In January 2017, he coauthors a Philips Research paper with Hadrien Bertrand, Roberto Ardon and Isabelle Bloch that classifies MRI images with deep networks tuned by random and Gaussian-process search.

  6. 2016–2024

    Head of AI at L'Oréal

    He joins L'Oréal in Paris in 2016 as head of its data science and machine learning lab, and he serves as head of AI there until 2024. In 2020, with Julien Despois and Frédéric Flament of L'Oréal Research & Innovation, he coauthors AgingMapGAN, which ages high-resolution face images using ethnicity-specific aging information and weak spatial supervision.

  7. 2020–2022

    Makeup transfer and virtual try-on

    In 2020, he coauthors CA-GAN with Robin Kips, Pietro Gori and Isabelle Bloch, a weakly supervised GAN that changes the color of lips or eyes for controllable makeup transfer. With L'Oréal, Modiface and Télécom Paris researchers, he presents an inverse graphics method for makeup synthesis from a reference image at a 2021 CVPR workshop. In 2022, the team maps a single example image into the parameters of a real-time augmented-reality renderer.

  8. Since December 2024

    Founding Evanova

    He founds Evanova in December 2024, after his years as head of AI at L'Oréal. Registered in April 2025, the company offers consulting, services and software development in artificial intelligence and computer vision, and he is its president.

Career

  1. Since 2024EvanovaPresident
  2. 2017–2016PhilipsResearch scientist
  3. 2016–2024L'OréalHead of AI
  4. 2014DailymotionData scientist
  5. 2014TwengaData scientist
  6. 2014–2016Assistance Publique – Hôpitaux de Paris
  7. 2011–2014Brain and Spine InstituteR&D manager
  8. 2009–2011DOSIsoftR&D engineer
  9. 2009–2014Direction de la Recherche Fondamentale
  10. 2008–2011Inserm
  11. 2008–2013Université Paris-Saclay
  12. 2008–2016Commissariat à l'Énergie Atomique et aux Énergies Alternatives
  13. 2006–2009CEA Paris-SaclayResearch engineer

Education

  1. 2009École normale supérieure de CachanPhD
  2. 2005École normale supérieure de CachanMaster's degree
  3. 2004Graduate School of Computer Science and Advanced Technologies

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Scikit-learn: Machine Learning in Python2012 · 69,582 citationsWith Alexandre Passos, Bertrand Thirion, David Cournapeau, Fabian Pedregosa, Matthieu Brucher and Olivier Grisel
  2. Cortical sulci recognition and spatial normalization2011 · Medical Image Analysis · 149 citations
  3. The effect of handedness on the shape of the central sulcus2011 · NeuroImage · 131 citations
  4. Anatomical correlations of the international 10–20 sensor placement system in infants2014 · NeuroImage · 111 citations
  5. Diffeomorphic Brain Registration Under Exhaustive Sulcal Constraints2011 · IEEE Transactions on Medical Imaging · 79 citations

From OpenAlex, as of 21 October 2026.

  1. 11% of AI doctorate holders from 2010 to 2016 founded or led a company or lab by 10 years after graduationIncludes Matthieu Perrot. We estimate how often AI doctorate holders founded or led a company or lab by 10 years after graduation, how waiting times compare between founders and top jobs, and which universities trained the most.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Commissariat à l'Énergie Atomique et aux Énergies Alternatives. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  3. 23% of people shaping AI who left a university for a company later returned to a universityIncludes Philips. We estimate that 468 of the 2,011 people shaping AI who left a university for a company (23%) later returned to a university after a median of 5 years, and 163 of 829 (20%) made the reverse round trip.
  4. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Université Paris-Saclay. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Matthieu Perrot yet. Send us a free-licensed photo.

  1. Matthieu PERROT (L'Oréal à PARIS) - Viadeo viadeo.journaldunet.com. Checked 20 October 2026.
  2. Matthieu Perrot | alphaXiv alphaxiv.org. Checked 20 October 2026.
  3. [1407.5602] Predictive support recovery with TV-Elastic Net penalty and logistic regression: an application to structural MRI arxiv.org. Checked 20 October 2026.
  4. Matthieu Perrot - Google Scholar scholar.google.com. Checked 20 October 2026.
  5. Scikit-learn: Machine Learning in Python jmlr.org. Checked 20 October 2026.
  6. Classification of MRI data using Deep Learning and Gaussian Process-based Model Selection arxiv.org. Checked 20 October 2026.
  7. AgingMapGAN (AMGAN): High-Resolution Controllable Face Aging with Spatially-Aware Conditional GANs arxiv.org. Checked 20 October 2026.
  8. CA-GAN: Weakly Supervised Color Aware GAN for Controllable Makeup Transfer arxiv.org. Checked 20 October 2026.
  9. Deep graphics encoder for real-time video makeup synthesis from example - Institut Polytechnique de Paris researchportal.ip-paris.fr. Checked 20 October 2026.
  10. [2205.06305] Real-time Virtual-Try-On from a Single Example Image through Deep Inverse Graphics and Learned Differentiable Renderers arxiv.org. Checked 20 October 2026.
  11. Evanova (91400) : siret, siren, TVA, adresse... entreprises.lefigaro.fr. Checked 20 October 2026.