Naive-Deep Face Recognition: Touching the Limit of LFW Benchmark or Not?

TL;DR

This paper examines the impact of large datasets on face recognition performance, achieving high accuracy on LFW, but highlights the gap between machine and human recognition in real-world scenarios.

Contribution

The paper presents a high-accuracy face recognition system trained on large data and discusses the challenges and differences between benchmark results and real-world applications.

Findings

Achieved 99.50% accuracy on LFW benchmark

Identified a gap between machine and human recognition performance

Outlined challenges and potential solutions for real-world face recognition

Abstract

Face recognition performance improves rapidly with the recent deep learning technique developing and underlying large training dataset accumulating. In this paper, we report our observations on how big data impacts the recognition performance. According to these observations, we build our Megvii Face Recognition System, which achieves 99.50% accuracy on the LFW benchmark, outperforming the previous state-of-the-art. Furthermore, we report the performance in a real-world security certification scenario. There still exists a clear gap between machine recognition and human performance. We summarize our experiments and present three challenges lying ahead in recent face recognition. And we indicate several possible solutions towards these challenges. We hope our work will stimulate the community's discussion of the difference between research benchmark and real-world applications.