EFaR 2023: Efficient Face Recognition Competition

Jan Niklas Kolf,Fadi Boutros,Jurek Elliesen,Markus Theuerkauf,Naser Damer,Mohamad Alansari,Oussama Abdul Hay,Sara Alansari,Sajid Javed,Naoufel Werghi,Klemen Grm,Vitomir Štruc,Fernando Alonso-Fernandez,Kevin Hernandez Diaz,Josef Bigun,Anjith George,Christophe Ecabert,Hatef Otroshi Shahreza,Ketan Kotwal,Sébastien Marcel,Iurii Medvedev,Bo Jin,Diogo Nunes,Ahmad Hassanpour,Pankaj Khatiwada,Aafan Ahmad Toor,Bian Yang

2023-08-08

Abstract:This paper presents the summary of the Efficient Face Recognition Competition (EFaR) held at the 2023 International Joint Conference on Biometrics (IJCB 2023). The competition received 17 submissions from 6 different teams. To drive further development of efficient face recognition models, the submitted solutions are ranked based on a weighted score of the achieved verification accuracies on a diverse set of benchmarks, as well as the deployability given by the number of floating-point operations and model size. The evaluation of submissions is extended to bias, cross-quality, and large-scale recognition benchmarks. Overall, the paper gives an overview of the achieved performance values of the submitted solutions as well as a diverse set of baselines. The submitted solutions use small, efficient network architectures to reduce the computational cost, some solutions apply model quantization. An outlook on possible techniques that are underrepresented in current solutions is given as well.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper aims to address the issue of efficient facial recognition models and to advance research in this field by organizing the Efficient Facial Recognition Competition (EFaR) at the 2023 International Joint Conference on Biometrics (IJCB 2023). Specifically, the goals of the paper include:

1. **Evaluating the latest efficient and lightweight facial recognition methods**: The paper summarizes various solutions submitted in the competition and evaluates them to identify the current best facial recognition models.

2. **Promoting the development of new technologies**: The competition encourages researchers to develop new techniques and methods to improve the efficiency and accuracy of facial recognition systems.

3. **Comprehensive evaluation of model performance and complexity**: Unlike previous competitions that focused solely on recognition accuracy, this competition also pays special attention to the model size (number of parameters) and the computational complexity required for inference (FLOPs), aiming to find models that achieve a good balance between accuracy and computational efficiency.

4. **Providing benchmarks and comparison standards**: The paper proposes a set of evaluation protocols and metrics that can be used for future research comparisons, thereby enhancing the comparability between different methods.

In summary, the main purpose of this paper is to evaluate and promote the development of efficient and lightweight facial recognition technologies, providing detailed experimental results analysis and future outlooks.