We introduce a high-performance fingerprint liveness feature extraction technique that secured first place in LivDet 2023 Fingerprint Representation Challenge. Additionally, we developed a practical fingerprint recognition system with 94.68% accuracy, earning second place in LivDet 2023 Liveness Detection in Action. By investigating various methods, particularly style transfer, we demonstrate improvements in accuracy and generalization when faced with limited training data. As a result, our approach achieved state-of-the-art performance in LivDet 2023 Challenges.