Pub Date : 2025-01-03DOI: 10.1109/TIFS.2025.3526063
Jun Feng;Yefan Wu;Hong Sun;Shunli Zhang;Debin Liu
Secure two-party neural network (2P-NN) inference allows the server with a neural network model and the client with inputs to perform neural network inference without revealing their private data to each other. However, the state-of-the-art 2P-NN inference still suffers from large computation and communication overhead especially when used in ImageNet-scale deep neural networks. In this work, we design and build Panther, a lightweight and efficient secure 2P-NN inference system, which has great efficiency in evaluating 2P-NN inference while safeguarding the privacy of the server and the client. At the core of Panther, we have new protocols for 2P-NN inference. Firstly, we propose a customized homomorphic encryption scheme to reduce burdensome polynomial multiplications in the homomorphic encryption arithmetic circuit of linear protocols. Secondly, we present a more efficient and communication concise design for the millionaires’ protocol, which enables non-linear protocols with less communication cost. Our evaluations over three sought-after varying-scale deep neural networks show that Panther outperforms the state-of-the-art 2P-NN inference systems in terms of end-to-end runtime and communication overhead. Panther achieves state-of-the-art performance with up to $24.95times $ speedup for linear protocols and $6.40 times $ speedup for non-linear protocols in WAN when compared to prior arts.
{"title":"Panther: Practical Secure Two-Party Neural Network Inference","authors":"Jun Feng;Yefan Wu;Hong Sun;Shunli Zhang;Debin Liu","doi":"10.1109/TIFS.2025.3526063","DOIUrl":"10.1109/TIFS.2025.3526063","url":null,"abstract":"Secure two-party neural network (2P-NN) inference allows the server with a neural network model and the client with inputs to perform neural network inference without revealing their private data to each other. However, the state-of-the-art 2P-NN inference still suffers from large computation and communication overhead especially when used in ImageNet-scale deep neural networks. In this work, we design and build Panther, a lightweight and efficient secure 2P-NN inference system, which has great efficiency in evaluating 2P-NN inference while safeguarding the privacy of the server and the client. At the core of Panther, we have new protocols for 2P-NN inference. Firstly, we propose a customized homomorphic encryption scheme to reduce burdensome polynomial multiplications in the homomorphic encryption arithmetic circuit of linear protocols. Secondly, we present a more efficient and communication concise design for the millionaires’ protocol, which enables non-linear protocols with less communication cost. Our evaluations over three sought-after varying-scale deep neural networks show that Panther outperforms the state-of-the-art 2P-NN inference systems in terms of end-to-end runtime and communication overhead. Panther achieves state-of-the-art performance with up to <inline-formula> <tex-math>$24.95times $ </tex-math></inline-formula> speedup for linear protocols and <inline-formula> <tex-math>$6.40 times $ </tex-math></inline-formula> speedup for non-linear protocols in WAN when compared to prior arts.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"1149-1162"},"PeriodicalIF":6.3,"publicationDate":"2025-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142924468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-30DOI: 10.1109/tifs.2024.3524160
Kai Liang, Songze Li, Ming Ding, Feng Tian, Youlong Wu
{"title":"Privacy-Preserving Coded Schemes for Multi-Server Federated Learning with Straggling Links","authors":"Kai Liang, Songze Li, Ming Ding, Feng Tian, Youlong Wu","doi":"10.1109/tifs.2024.3524160","DOIUrl":"https://doi.org/10.1109/tifs.2024.3524160","url":null,"abstract":"","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 1","pages":""},"PeriodicalIF":6.8,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142905582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Specific emitter identification (SEI) is crucial in the Internet of Everything (IoE). Over the past decade, deep learning (DL) and broad learning (BL)-enabled SEI technologies have emerged. Both DL- and BL-based SEI methods rely on extensive radio frequency (RF) signal samples and corresponding labels, but labeling unknown signals is a considerable overhead and costly task. Consequently, many researchers have begun exploring semi-supervised learning techniques to address the semi-supervised SEI (SS-SEI) problem with limited labeled RF signals. However, existing SS-SEI solutions often prioritize identification performance, leading to high computational overheads and lacking iterability and scalability. To overcome these challenges, this paper proposes a novel SS-SEI solution, termed deep cloud and broad edge (DCBE). This approach integrates a DL-based SEI method at the cloud server with an updatable BL-based SEI method at the edge node. Initially, several DL-based SEI models are trained using labeled historical data at the cloud server. Meanwhile, an updatable BL-based SEI method is deployed locally on the edge node to identify unlabelled signals. When the DCBE solution is operational, edge nodes capture real-time unlabelled RF signals. The pre-trained DL-based SEI method and the locally BL-based SEI method jointly identify these RF signals. The identification results, along with the new real-time RF signals, are then used to update the weights of the BL-based SEI method at the edge nodes. The DCBE SS-SEI solution is validated using an open-source, large-scale, real-world automatic dependent surveillance-broadcast (ADS-B) dataset. Experimental results demonstrate that the proposed DCBE solution offers significant advantages in terms of SS-SEI performance, reduced computational overhead without GPU dependency, and system robustness in complex environments.
{"title":"Enhancing Specific Emitter Identification: A Semi-Supervised Approach With Deep Cloud and Broad Edge Integration","authors":"Yibin Zhang;Yuchao Liu;Juzhen Wang;Qi Xuan;Yun Lin;Guan Gui","doi":"10.1109/TIFS.2024.3524157","DOIUrl":"10.1109/TIFS.2024.3524157","url":null,"abstract":"Specific emitter identification (SEI) is crucial in the Internet of Everything (IoE). Over the past decade, deep learning (DL) and broad learning (BL)-enabled SEI technologies have emerged. Both DL- and BL-based SEI methods rely on extensive radio frequency (RF) signal samples and corresponding labels, but labeling unknown signals is a considerable overhead and costly task. Consequently, many researchers have begun exploring semi-supervised learning techniques to address the semi-supervised SEI (SS-SEI) problem with limited labeled RF signals. However, existing SS-SEI solutions often prioritize identification performance, leading to high computational overheads and lacking iterability and scalability. To overcome these challenges, this paper proposes a novel SS-SEI solution, termed deep cloud and broad edge (DCBE). This approach integrates a DL-based SEI method at the cloud server with an updatable BL-based SEI method at the edge node. Initially, several DL-based SEI models are trained using labeled historical data at the cloud server. Meanwhile, an updatable BL-based SEI method is deployed locally on the edge node to identify unlabelled signals. When the DCBE solution is operational, edge nodes capture real-time unlabelled RF signals. The pre-trained DL-based SEI method and the locally BL-based SEI method jointly identify these RF signals. The identification results, along with the new real-time RF signals, are then used to update the weights of the BL-based SEI method at the edge nodes. The DCBE SS-SEI solution is validated using an open-source, large-scale, real-world automatic dependent surveillance-broadcast (ADS-B) dataset. Experimental results demonstrate that the proposed DCBE solution offers significant advantages in terms of SS-SEI performance, reduced computational overhead without GPU dependency, and system robustness in complex environments.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"1092-1105"},"PeriodicalIF":6.3,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142905262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-27DOI: 10.1109/tifs.2024.3523765
Chenhao Wang, Yang Ming, Hang Liu, Yutong Deng, Yi Zhao, Songnian Zhang
{"title":"Security-Enhanced Data Transmission with Fine-Grained and Flexible Revocation for DTWNs","authors":"Chenhao Wang, Yang Ming, Hang Liu, Yutong Deng, Yi Zhao, Songnian Zhang","doi":"10.1109/tifs.2024.3523765","DOIUrl":"https://doi.org/10.1109/tifs.2024.3523765","url":null,"abstract":"","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"3 1","pages":""},"PeriodicalIF":6.8,"publicationDate":"2024-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Blockchain-empowered Keyword Searchable Provable Data Possession for Large Similar Data","authors":"Ying Miao, Keke Gai, Jing Yu, Yu’an Tan, Liehuang Zhu, Weizhi Meng","doi":"10.1109/tifs.2024.3516563","DOIUrl":"https://doi.org/10.1109/tifs.2024.3516563","url":null,"abstract":"","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"114 1","pages":""},"PeriodicalIF":6.8,"publicationDate":"2024-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reconnaissance-Strike Complex: A Network-layer Solution to the Natural Forking in Blockchain","authors":"Anlin Chen, Shengling Wang, Hongwei Shi, Yu Guo, Xiuzhen Cheng","doi":"10.1109/tifs.2024.3523767","DOIUrl":"https://doi.org/10.1109/tifs.2024.3523767","url":null,"abstract":"","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"151 1","pages":""},"PeriodicalIF":6.8,"publicationDate":"2024-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Gradient inversion attacks (GIAs) pose significant challenges to the privacy-preserving paradigm of distributed learning. These attacks employ carefully designed strategies to reconstruct victim’s private training data from their shared gradients. However, existing work mainly focuses on attacks and defenses for image-modal data, while the study for text-modal data remains scarce. Furthermore, the performance of the limited attack researches on text-modal data is also unsatisfactory, which can be partially attributed to the finer granularity of text data compared to image. To bridge the existing research gap, we propose a high-fidelity attack method tailored for Transformer-based language models (LMs). In our method, we initially reconstruct the label space of the victim’s training data by leveraging the characteristics of the Transformer architecture. After that, we propose a shallow-to-deep paradigm to facilitate gradient matching, which can significantly improve the attack performance. Furthermore, we develop a weighted surrogate loss that resolves the consistent deviation issue present in current attack researches. A substantial number of experiments on Transformer-based LMs (e.g., Bert and GPT) demonstrate that our attack is competitive and significantly outperforms existing methods. In the final part of this paper, we investigate the influence of the inherent position embedding module within the Transformer architecture on attack performance, and based on the analysis results, we propose a countermeasure to alleviate part of the privacy leakage issue in distributed learning.
{"title":"Gradient Inversion of Text-Modal Data in Distributed Learning","authors":"Zipeng Ye;Wenjian Luo;Qi Zhou;Yubo Tang;Zhenqian Zhu;Yuhui Shi;Yan Jia","doi":"10.1109/TIFS.2024.3522792","DOIUrl":"10.1109/TIFS.2024.3522792","url":null,"abstract":"Gradient inversion attacks (GIAs) pose significant challenges to the privacy-preserving paradigm of distributed learning. These attacks employ carefully designed strategies to reconstruct victim’s private training data from their shared gradients. However, existing work mainly focuses on attacks and defenses for image-modal data, while the study for text-modal data remains scarce. Furthermore, the performance of the limited attack researches on text-modal data is also unsatisfactory, which can be partially attributed to the finer granularity of text data compared to image. To bridge the existing research gap, we propose a high-fidelity attack method tailored for Transformer-based language models (LMs). In our method, we initially reconstruct the label space of the victim’s training data by leveraging the characteristics of the Transformer architecture. After that, we propose a shallow-to-deep paradigm to facilitate gradient matching, which can significantly improve the attack performance. Furthermore, we develop a weighted surrogate loss that resolves the consistent deviation issue present in current attack researches. A substantial number of experiments on Transformer-based LMs (e.g., Bert and GPT) demonstrate that our attack is competitive and significantly outperforms existing methods. In the final part of this paper, we investigate the influence of the inherent position embedding module within the Transformer architecture on attack performance, and based on the analysis results, we propose a countermeasure to alleviate part of the privacy leakage issue in distributed learning.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"928-943"},"PeriodicalIF":6.3,"publicationDate":"2024-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-25DOI: 10.1109/TIFS.2024.3522758
Zhenxin Cai;Yu Wang;Guan Gui;Jin Sha
Radio frequency fingerprint identification (RFFI) is regarded as one of the most promising techniques for managing and regulating Internet of Things (IoT) devices. This technology analyzes the unique electromagnetic signals emitted by wireless devices to enable precise identification and authentication. Most existing RFFI methods focus on RF signals collected in specific scenarios. However, in real-world applications, signals are often collected at different times or from varying deployment locations, leading to differences between the training and testing distributions. The study of RFFI methods under these conditions remains underexplored. To address this gap, this paper introduces a cross-domain RFFI framework centered on adaptive semantic augmentation (ASA). The framework integrates a computationally efficient multi-resolution spectrogram decomposition strategy with a feature-sensitive multi-scale network. The ASA method enhances RFFI accuracy in cross-domain settings by linearly interpolating between two distinct semantic features to create new semantics for further identification. The proposed approach leverages two-dimensional discrete wavelet transform (2D-DWT) to decompose the raw spectrogram into four sub-bands, followed by a multi-scale network to extract critical semantic features for the ASA method. Simulation results show that the proposed ASA method significantly improves Unmanned Aerial Vehicle (UAV) identification performance, achieving accuracies of 93.05% and 98.90% on two different cross-domain datasets, respectively, outperforming existing data augmentation (DA) methods. Furthermore, generalizability validation demonstrates that the proposed method performs outstandingly across other Internet of Things (IoT) applications.
{"title":"Toward Robust Radio Frequency Fingerprint Identification via Adaptive Semantic Augmentation","authors":"Zhenxin Cai;Yu Wang;Guan Gui;Jin Sha","doi":"10.1109/TIFS.2024.3522758","DOIUrl":"10.1109/TIFS.2024.3522758","url":null,"abstract":"Radio frequency fingerprint identification (RFFI) is regarded as one of the most promising techniques for managing and regulating Internet of Things (IoT) devices. This technology analyzes the unique electromagnetic signals emitted by wireless devices to enable precise identification and authentication. Most existing RFFI methods focus on RF signals collected in specific scenarios. However, in real-world applications, signals are often collected at different times or from varying deployment locations, leading to differences between the training and testing distributions. The study of RFFI methods under these conditions remains underexplored. To address this gap, this paper introduces a cross-domain RFFI framework centered on adaptive semantic augmentation (ASA). The framework integrates a computationally efficient multi-resolution spectrogram decomposition strategy with a feature-sensitive multi-scale network. The ASA method enhances RFFI accuracy in cross-domain settings by linearly interpolating between two distinct semantic features to create new semantics for further identification. The proposed approach leverages two-dimensional discrete wavelet transform (2D-DWT) to decompose the raw spectrogram into four sub-bands, followed by a multi-scale network to extract critical semantic features for the ASA method. Simulation results show that the proposed ASA method significantly improves Unmanned Aerial Vehicle (UAV) identification performance, achieving accuracies of 93.05% and 98.90% on two different cross-domain datasets, respectively, outperforming existing data augmentation (DA) methods. Furthermore, generalizability validation demonstrates that the proposed method performs outstandingly across other Internet of Things (IoT) applications.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"1037-1048"},"PeriodicalIF":6.3,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-25DOI: 10.1109/TIFS.2024.3522770
Jiancun Wu;Engang Tian;Chen Peng;Zhiru Cao
This paper is concerned with the security issues related to integrated attack-defense strategy for a category of multi-sensor networked control systems with state saturation constraints. In general, existing denial-of-service (DoS) attack models typically conduct indiscriminate attacks on data packets, disregarding the significance of the attacked data packets to the system. Note that the measurement data from different sensor nodes possesses varying levels of importance. In light of this, we first propose a novel form of attack from the perspective of attack design, known as a data-importance-aware attack. The importance of data refers to the quantitative impact of the measured values at each sensor node on the stable and safe operation of the entire system. As such, the proposed attack has the awareness to launch attacks against critical sensor nodes, rendering data unable to be transmitted. Then, an attack-node-dependent security controller is devised from the defender’s perspective against the constructed attack, which can effectively resist the impact of attacks and stabilize the system. By employing the Lyapunov functional method, sufficient conditions are derived to ensure the asymptotic stability of the closed-loop system. Finally, the reliability and effectiveness of the node importance-aware attack strategy and control countermeasure are validated by numerical simulation.
{"title":"Data-Importance-Aware Attack Strategy Design and Secure Control Countermeasure","authors":"Jiancun Wu;Engang Tian;Chen Peng;Zhiru Cao","doi":"10.1109/TIFS.2024.3522770","DOIUrl":"10.1109/TIFS.2024.3522770","url":null,"abstract":"This paper is concerned with the security issues related to integrated attack-defense strategy for a category of multi-sensor networked control systems with state saturation constraints. In general, existing denial-of-service (DoS) attack models typically conduct indiscriminate attacks on data packets, disregarding the significance of the attacked data packets to the system. Note that the measurement data from different sensor nodes possesses varying levels of importance. In light of this, we first propose a novel form of attack from the perspective of attack design, known as a data-importance-aware attack. The importance of data refers to the quantitative impact of the measured values at each sensor node on the stable and safe operation of the entire system. As such, the proposed attack has the awareness to launch attacks against critical sensor nodes, rendering data unable to be transmitted. Then, an attack-node-dependent security controller is devised from the defender’s perspective against the constructed attack, which can effectively resist the impact of attacks and stabilize the system. By employing the Lyapunov functional method, sufficient conditions are derived to ensure the asymptotic stability of the closed-loop system. Finally, the reliability and effectiveness of the node importance-aware attack strategy and control countermeasure are validated by numerical simulation.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"944-954"},"PeriodicalIF":6.3,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-25DOI: 10.1109/TIFS.2024.3521323
Yu Tian;Kunbo Zhang;Yalin Huang;Leyuan Wang;Yue Liu;Zhenan Sun
Despite the development of spectral sensors and spectral data-driven learning methods which have led to significant advances in face anti-spoofing (FAS), the singular dimensionality of spectral information often results in poor robustness and weak generalization. Polarization, another fundamental property of light, can reveal intrinsic differences between genuine and fake faces with advantaged performance in precision, robustness, and generalizability. In this paper, we propose a facial image translation method from visible light (VIS) to polarization (VPT), capable of generating valuable polarimetric optical characteristics for facial presentation attack detection using VIS spectrum information input only. Specifically, the VPT method adopts a multi-stream network structure, comprising a main network and two branch networks, to translate VIS images into degree of polarization (DoP) images and Stokes polarization parameters ${S}_{1}$ and ${S}_{2}$ . To further improve image translation quality, we introduce a frequency-domain consistency loss as a complement to the existing spatial losses to narrow the gap in the frequency domain. The physical mapping relations for the DoP and Stokes parameters are employed, and the Stokes loss is designed to ensure that the generated polarization modalities conform to objective physical laws. Extensive experiments on the CASIA-Polar and CASIA-SURF datasets demonstrate the superiority of VPT over other baseline methods in terms of polarization image quality and its remarkable performance in the FAS task. This work leverages the inherent physical advantages of polarization information in material discrimination tasks while addressing hardware limitations in polarization image collection, proposing a novel solution for face recognition system security control.
{"title":"Cross-Optical Property Image Translation for Face Anti-Spoofing: From Visible to Polarization","authors":"Yu Tian;Kunbo Zhang;Yalin Huang;Leyuan Wang;Yue Liu;Zhenan Sun","doi":"10.1109/TIFS.2024.3521323","DOIUrl":"10.1109/TIFS.2024.3521323","url":null,"abstract":"Despite the development of spectral sensors and spectral data-driven learning methods which have led to significant advances in face anti-spoofing (FAS), the singular dimensionality of spectral information often results in poor robustness and weak generalization. Polarization, another fundamental property of light, can reveal intrinsic differences between genuine and fake faces with advantaged performance in precision, robustness, and generalizability. In this paper, we propose a facial image translation method from visible light (VIS) to polarization (VPT), capable of generating valuable polarimetric optical characteristics for facial presentation attack detection using VIS spectrum information input only. Specifically, the VPT method adopts a multi-stream network structure, comprising a main network and two branch networks, to translate VIS images into degree of polarization (DoP) images and Stokes polarization parameters <inline-formula> <tex-math>${S}_{1}$ </tex-math></inline-formula> and <inline-formula> <tex-math>${S}_{2}$ </tex-math></inline-formula>. To further improve image translation quality, we introduce a frequency-domain consistency loss as a complement to the existing spatial losses to narrow the gap in the frequency domain. The physical mapping relations for the DoP and Stokes parameters are employed, and the Stokes loss is designed to ensure that the generated polarization modalities conform to objective physical laws. Extensive experiments on the CASIA-Polar and CASIA-SURF datasets demonstrate the superiority of VPT over other baseline methods in terms of polarization image quality and its remarkable performance in the FAS task. This work leverages the inherent physical advantages of polarization information in material discrimination tasks while addressing hardware limitations in polarization image collection, proposing a novel solution for face recognition system security control.","PeriodicalId":13492,"journal":{"name":"IEEE Transactions on Information Forensics and Security","volume":"20 ","pages":"1192-1205"},"PeriodicalIF":6.3,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10816165","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：