Jan Niklas Kolf
JK
01
About

Hi, I'm Jan Niklas Kolf

I am a research scientist at Fraunhofer IGD and a doctoral researcher at TU Darmstadt. My PhD is supervised by Arjan Kuijper, Naser Damer, and Fadi Boutros, and is conducted within the Embedded Biometrics project of ATHENE, the German National Research Center for Applied Cybersecurity. My work focuses on advancing biometric recognition systems from both algorithmic and deployment-oriented perspectives.

I received the European Biometrics Research Award and the Best Presentation Award at the 19th European Biometrics Max Snijder, Research, and Industry Awards of the European Association for Biometrics for my doctoral research.

In summer 2025, I was a research intern at Meta Reality Labs in Burlingame, California, under the supervision of Ozan Tezcan.

I previously completed master’s degrees in Autonomous Systems and Visual Computing at TU Darmstadt, graduating with honors in both programs.

Research profile

Research across face recognition, knowledge distillation, model quantization, face image quality assessment, and synthetic data from deep generative models.

Portrait
Portrait
02
Publications

Selected Publications

Google Scholar

  1. BLenDeR: Blended Text Embeddings and Diffusion Residuals for Intra-Class Image Synthesis in Deep Metric Learning

    Jan Niklas Kolf, Ozan Tezcan, Justin Theiss, Hyung Jun Kim, Wentao Bao, Bhargav Bhushanam, Khushi Gupta, Arun Kejariwal, Naser Damer, Fadi Boutros

    Under review

    arXiv Code Hugging Face

  2. PreFIQs: Face Image Quality Is What Survives Pruning

    Jan Niklas Kolf, Guray Ozgur, Andrea Atzori, Žiga Babnik, Vitomir Štruc, Naser Damer, Fadi Boutros

    CVPR 2026 Workshops

    arXiv PDF Code

  3. GraFIQs: Face Image Quality Assessment Using Gradient Magnitudes

    Jan Niklas Kolf, Naser Damer, Fadi Boutros

    CVPR 2024 Workshops

    arXiv PDF Code
03
Research

Current work

  • a. Deep Generative Models for privacy-preserving Face Recognition
  • b. Training- and label-free Face Image Quality Assessment
  • c. Knowledge Distillation for efficient Face Recognition
04
Contact

Send a note with context.

Google Scholar ORCID