Efficient Low-Resolution Face Recognition via Bridge Distillation

TL;DR

This paper introduces a bridge distillation method to efficiently adapt high-resolution face recognition models for low-resolution images, achieving high speed and accuracy with a lightweight model.

Contribution

It proposes a two-step distillation process that transfers knowledge from high-resolution to low-resolution face recognition models, enabling efficient deployment.

Findings

Model with 0.21M parameters achieves high accuracy on low-resolution faces.

The approach significantly improves inference speed on various devices.

The method outperforms existing lightweight face recognition models.

Abstract

Face recognition in the wild is now advancing towards light-weight models, fast inference speed and resolution-adapted capability. In this paper, we propose a bridge distillation approach to turn a complex face model pretrained on private high-resolution faces into a light-weight one for low-resolution face recognition. In our approach, such a cross-dataset resolution-adapted knowledge transfer problem is solved via two-step distillation. In the first step, we conduct cross-dataset distillation to transfer the prior knowledge from private high-resolution faces to public high-resolution faces and generate compact and discriminative features. In the second step, the resolution-adapted distillation is conducted to further transfer the prior knowledge to synthetic low-resolution faces via multi-task learning. By learning low-resolution face representations and mimicking the adapted high-resolution knowledge, a light-weight student model can be constructed with high efficiency and promising accuracy in recognizing low-resolution faces. Experimental results show that the student model performs impressively in recognizing low-resolution faces with only 0.21M parameters and 0.057MB memory. Meanwhile, its speed reaches up to 14,705, ~934 and 763 faces per second on GPU, CPU and mobile phone, respectively.