Perceptual hash functions
Security and robustness of systems used to recognize visually similar media, including PhotoDNA, NeuralHash and PDQ.
Research profile
Cryptanalysis with practical and societal impact
Time-memory trade-offs
Attack models, variants and implementation aspects of TMTO techniques, building on my PhD work on Rainbow tables.
Security protocols
Authentication and password-manager protocols, with a focus on privacy, integrity and realistic adversarial models.
Featured project
PhotoDNA and perceptual hash functions
Recent work studies the reliability of deployed perceptual hash functions, their collision behavior, and the consequences for large-scale content scanning.
Selected publications
Recent research
Breaking Widely Deployed Perceptual Hash Functions: Black-Box Collisions in Apple NeuralHash and Microsoft PhotoDNA
Diane Leblanc-Albarel and Bart Preneel. 35th USENIX Security Symposium, 2026.
White-Box Attacks on PhotoDNA Perceptual Hash Function
Maxime Deryck, Diane Leblanc-Albarel and Bart Preneel. Preprint.
In the Vault, But Not Safe: Exploring the Threat of Covert Password Manager Providers
Gildas Avoine, Xavier Carpent and Diane Leblanc-Albarel. Preprint.
Black-box Collision Attacks on the NeuralHash Perceptual Hash Function
Diane Leblanc-Albarel and Bart Preneel. Preprint.
Time-Memory Trade-Offs Sound the Death Knell for GPRS and GSM
Work on TMTO cryptanalysis and legacy mobile communication systems.
Slides and seminars
Talks
AFSIN · September 2026
From detection to reporting: online CSAM detection procedures, mechanisms and limits
French slides on reporting pipelines, platform scanning, NCMEC, EU/FR regulation and PhotoDNA attacks.
Slides in French
SoSySec seminar · Rennes · June 2026
Opening Pandora’s Box: White-Box Attacks on Microsoft’s PhotoDNA Perceptual Hash Function
Seminar slides introducing Alleged PhotoDNA, white-box attacks, and client-side scanning implications.
Slides Event pageIn the press
Other works and events
I co-authored the French book chapter Les mots de passe à bout de souffle with Gildas Avoine in Le Calcul à Découvert, and the article Comment choisir un bon mot de passe ? in The Conversation.
I also contribute to inclusion activities, including an IEEE Women in Engineering book circle at KU Leuven and past Python workshops for middle-school girls in Rennes.
Previous positions
I completed my PhD in October 2023 under the supervision of Gildas Avoine in the SPICY team at IRISA, Rennes. My PhD focused on Time-Memory Trade-Off attacks, especially Rainbow tables.
Before my PhD, I graduated in Mathematical Engineering from INSA Rouen Normandie and completed a Master 2 in Information Systems Security at the University of Rouen Normandie.