RVFNet: A Network Model for Keystroke Authentication

doi:10.11871/jfdc.issn.2096-742X.2025.06.002

Abstract

Abstract:

[Objective] This paper aims to address the accuracy deficit of traditional keystroke-based authentication caused by overlooking global behavioral patterns. [Methods] First, we convert keystroke sequences into visual representations via three-channel image encoding, leveraging CNNs’ proven image-recognition power to build precise user-identity models. Second, we introduce the novel RVFNet architecture, integrating our proprietary Feature Enhancement Block (FEBlock) to deeply fuse the spatiotemporal dynamics of keystroke behavior. [Results] On the Aalto desktop dataset, RVFNet attains an AUC of 98.62% and an EER as low as 1.46% with only 50 keystrokes. [Limitations] The model’s large parameter count limits deployment on resource-constrained devices. [Conclusions] Experimental results confirm the method’s effectiveness in detection accuracy.

Key words: information security, authentication, biometrics, keystroke dynamics, neural network

WANG Chang,LI Peiyu,LIN Yusong. RVFNet: A Network Model for Keystroke Authentication[J]. Frontiers of Data and Computing, 2025, 7(6): 13-22, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2025.06.002.

Figures/Tables 15

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Table 1

Table 2

Fig.7

Table 3

Fig.8

Fig.9

Table 4

Fig.10

Table 5

References 22

[1]	郑起鼎, 王贺祥, 张洪玮. 一种基于分布式密钥与BLS签名的跨域认证方案[J]. 数据与计算发展前沿(中英文), 2024, 6(5): 13-23.
[2]	荆继武, 龙春, 李畅. 网络安全技术的新趋势探讨[J]. 数据与计算发展前沿, 2021, 3(3): 1-8.
[3]	潘静, 李博逊, 万贝琳, 等. 基于格的零知识证明综述[J]. 信息网络安全, 2025, 25(5): 679-688.
[4]	MAHARJAN P, SHRESTHA K, BHATTA T, et al. Keystroke dynamics based hybrid nanogenerators for biometric authentication and identification using artificial intelligence[J]. Advanced Science, 2021, 8(15): 2100711.
[5]	BUDZYS A, KURASOVA O, MEDVEDEV V. Integrating deep learning and data fusion for advanced keystroke dynamics authentication[J]. Computer Standards & Interfaces, 2025, 92: 103931.
[6]	MONROSE F, RUBIN A. Authentication via keystroke dynamics[C]// Proceedings of the 4th ACM Conference on Computer and Communications Security, 1997: 48-56.
[7]	GUNETTI D, PICARDI C. Keystroke analysis of free text[J]. ACM Transactions on Information and System Security (TISSEC), 2005, 8(3): 312-347.
[8]	LU X, ZHANG S, HUI P, et al. Continuous authentication by free-text keystroke based on CNN and RNN[J]. Computers & Security, 2020, 96: 101861.
[9]	YANG H, SUN D, WANG Y, et al. Fktan: Fusion keystroke time-textual attention networks for continuous authentication[C]// 2021 IEEE Symposium on Computers and Communications (ISCC), IEEE, 2021: 1-6.
[10]	BUDZYS A, KURASOVA O, MEDVEDEV V. Deep learning-based authentication for insider threat detection in critical infrastructure[J]. Artificial Intelligence Review, 2024, 57(10): 272.
[11]	ACIEN A, MORALES A, MONACO J V, et al. TypeNet: Deep learning keystroke biometrics[J]. IEEE Transactions on Biometrics, Behavior, and Identity Science, 2021, 4(1): 57-70.
[12]	STRAGAPEDE G, DELGADO-SANTOS P, TOLOSANA R, et al. TypeFormer: Transformers for mobile keystroke biometrics[J]. Neural Computing and Applications, 2024, 36(29): 18531-18545.
[13]	TEH P S, TEOH A B J, YUE S. A survey of keystroke dynamics biometrics[J]. The Scientific World Journal, 2013, 2013(1): 408280.
[14]	HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]// Proceedings of the IEEE conference on computer vision and pattern recognition, 2016: 770-778.
[15]	RAHMAN N A, YUSOFF N. Modulated spike-time dependent plasticity (STDP)-based learning for spiking neural network (SNN): A review[J]. Neurocomputing, 2025, 618: 129170.
[16]	HAN K, WANG Y, CHEN H, et al. A survey on vision transformer[J]. IEEE transactions on pattern analysis and machine intelligence, 2022, 45(1): 87-110.
[17]	WOO S, PARK J, LEE J Y, et al. Cbam: Convolutional block attention module[C]// Proceedings of the European conference on computer vision (ECCV), 2018: 3-19.
[18]	DHAKAL V, FEIT A M, KRISTENSSON P O, et al. Observations on typing from 136 million keystrokes[C]// Proceedings of the 2018 CHI conference on human factors in computing systems, 2018: 1-12.
[19]	HUANG J, LING C X. Using AUC and accuracy in evaluating learning algorithms[J]. IEEE Transactions on knowledge and Data Engineering, 2005, 17(3): 299-310.
[20]	ISMAIL M G, SALEM M A M, ABDELGHANY M A, et al. Outlier detection for keystroke biometric user authentication[J]. PeerJ Computer Science, 2024, 10: e2086.
[21]	YANG L, LI C, YOU R, et al. TKCA: a timely keystroke-based continuous user authentication with short keystroke sequence in uncontrolled settings[J]. Cybersecurity, 2021, 4: 1-16.
[22]	HAZAN I, MARGALIT O, ROKACH L. Supporting unknown number of users in keystroke dynamics models[J]. Knowledge-Based Systems, 2021, 221: 106982.

击键序列长度L	TypeNet	RVFNet (我们的)
30	11.5%	8.11%
50	8.8%	1.46%
70	6%	0.93%
100	5.7%	4.8%

方法	AUC	EER
CNN+RNN^[8]	—	5.97%
FKTAN^[9]	—	4.78%
TypeFormer^[12]	—	3.25%
HBOS^[20]	97.79%	5.97%
TKCA^[21]	—	2.78%
X-means with QT^[22]	94.2%	11.2%
RVFNet (我们的)	98.62%	1.46%

击键序列长度L	AUC	EER
30	92.5%	8.11%
50	98.62%	1.46%
70	99.15%	0.93%
100	95.2%	4.8%

击键特征	AUC	EER
其他研究	97.5%	3.17%
本研究	99.15%	0.93%

方法	AUC	EER
×FEBlock	96.8%	3.72%
×CBAM	98.4%	1.6%
RVFNet	99.15%	0.93%