LivDet – Liveness Detection Competitions

Spoof attacks also known as presentation attacks use artificial replicas of biometric traits (e.g., masks, printed images) to fool a recognition system. Liveness detection methods work by distinguishing genuine, live biometrics from these fakes.

The goal of the LivDet Competitions is to benchmark and compare liveness detection approaches under a common protocol and large-scale datasets of live and spoof samples. Open to both academic and industrial teams, LivDet evaluates software-only algorithms as well as fully integrated hardware/software systems.

Since its debut in 2009, LivDet has run in 2009, 2011, 2013, 2015, 2017, 2019, 2020, 2021, 2023, 2024, and 2025 providing invaluable insights into advances in detection techniques.

Today, LivDet covers three biometric modalities iris, fingerprint, and face and plans are underway to extend the competition to additional traits in future editions.

Each LivDet edition features two tracks:

• Part I: Software-based — Evaluates algorithms on a shared live/spoof dataset.
• Part II: System-based — Tests complete hardware/software solutions using live and spoof samples.

After each competition, anonymized live and spoof datasets are released to the research community upon request.