RPG-Palm: Realistic Pseudo-data Generation for Palmprint Recognition

TL;DR

This paper introduces RPG-Palm, a novel pseudo-data generation model for palmprint recognition that synthesizes realistic palmprints to enhance recognition performance and reduce dependence on large real datasets.

Contribution

The paper presents a new generative model with a conditional modulation generator and identity-aware loss to produce diverse, identity-consistent palmprints, advancing palmprint recognition research.

Findings

Synthetic palmprints significantly improve recognition accuracy.

The model outperforms state-of-the-art methods by over 5% and 14%.

Effective even with only 10% real training data.

Abstract

Palmprint recently shows great potential in recognition applications as it is a privacy-friendly and stable biometric. However, the lack of large-scale public palmprint datasets limits further research and development of palmprint recognition. In this paper, we propose a novel realistic pseudo-palmprint generation (RPG) model to synthesize palmprints with massive identities. We first introduce a conditional modulation generator to improve the intra-class diversity. Then an identity-aware loss is proposed to ensure identity consistency against unpaired training. We further improve the B\'ezier palm creases generation strategy to guarantee identity independence. Extensive experimental results demonstrate that synthetic pretraining significantly boosts the recognition model performance. For example, our model improves the state-of-the-art B\'ezierPalm by more than $5\%$ and $14\%$ in terms of TAR@FAR=1e-6 under the $1:1$ and $1:3$ Open-set protocol. When accessing only $10\%$ of the real training data, our method still outperforms ArcFace with $100\%$ real training data, indicating that we are closer to real-data-free palmprint recognition.