Publications

2025

1. Type2Branch: Keystroke Biometrics based on a Dual-branch Architecture with Attention Mechanisms and Set2set Loss

   Nahuel González, Giuseppe Stragapede, Rubén Vera-Rodriguez, and 1 more author

   Under review, 2025

   Abs arXiv Bib Code

   In 2021, the pioneering work on TypeNet showed that keystroke dynamics verification could scale to hundreds of thousands of users with minimal performance degradation. Recently, the KVC-onGoing competition has provided an open and robust experimental protocol for evaluating keystroke dynamics verification systems of such scale, including considerations of algorithmic fairness. This article describes Type2Branch, the model and techniques that achieved the lowest error rates at the KVC-onGoing, in both desktop and mobile scenarios. The novelty aspects of the proposed Type2Branch include: i) synthesized timing features emphasizing user behavior deviation from the general population, ii) a dual-branch architecture combining recurrent and convolutional paths with various attention mechanisms, iii) a new loss function named Set2set that captures the global structure of the embedding space, and iv) a training curriculum of increasing difficulty. Considering five enrollment samples per subject of approximately 50 characters typed, the proposed Type2Branch achieves state-of-the-art performance with mean per-subject EERs of 0.77% and 1.03% on evaluation sets of respectively 15,000 and 5,000 subjects for desktop and mobile scenarios. With a uniform global threshold for all subjects, the EERs are 3.25% for desktop and 3.61% for mobile, outperforming previous approaches by a significant margin.

   @article{type2branch,
     title = {Type2Branch: Keystroke Biometrics based on a Dual-branch Architecture with Attention Mechanisms and Set2set Loss},
     author = {Gonz{\'a}lez, Nahuel and Stragapede, Giuseppe and Vera-Rodriguez, Rub{\'e}n and Tolosana, Rub{\'e}n},
     journal = {Under review},
     year = {2025},
   }

2. KVC-onGoing: Keystroke Verification Challenge

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 27 more authors

   Pattern Recognition, 2025

   Abs DOI arXiv Bib Supp Code Website

   This article presents the Keystroke Verification Challenge - onGoing (KVC-onGoing), on which researchers can easily benchmark their systems in a common platform using large-scale public databases, the Aalto University Keystroke databases, and a standard experimental protocol. The keystroke data consist of tweet-long sequences of variable transcript text from over 185,000 subjects, acquired through desktop and mobile keyboards simulating real-life conditions. The results on the evaluation set of KVC-onGoing have proved the high discriminative power of keystroke dynamics, reaching values as low as 3.33% of Equal Error Rate (EER) and 11.96% of False Non-Match Rate (FNMR) @1% False Match Rate (FMR) in the desktop scenario, and 3.61% of EER and 17.44% of FNMR @1% at FMR in the mobile scenario, significantly improving previous state-of-the-art results. Concerning demographic fairness, the analyzed scores reflect the subjects’ age and gender to various extents, not negligible in a few cases. The framework runs on CodaLab.

   @article{stragapede2025kvc,
     title = {KVC-onGoing: Keystroke Verification Challenge},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and DeAndres-Tame, Ivan and Damer, Naser and Fierrez, Julian and Ortega-Garcia, Javier and Acien, Alejandro and Gonzalez, Nahuel and Shadrikov, Andrei and Gordin, Dmitrii and Schmitt, Leon and Wimmer, Daniel and Großmann, Christoph and Krieger, Joerdis and Heinz, Florian and Krestel, Ron and Mayer, Christoffer and Haberl, Simon and Gschrey, Helena and Yamagishi, Yosuke and Saha, Sanjay and Rasnayaka, Sanka and Wickramanayake, Sandareka and Sim, Terence and Gutfeter, Weronika and Baran, Adam and Krzysztoń, Mateusz and Jaskóła, Przemysław},
     journal = {Pattern Recognition},
     volume = {161},
     pages = {111287},
     year = {2025},
     publisher = {Elsevier},
     doi = {https://doi.org/10.1016/j.patcog.2024.111287},
   }

2024

1. TypeFormer: Transformers for mobile keystroke biometrics

   Giuseppe Stragapede, Paula Delgado-Santos, Ruben Tolosana, and 3 more authors

   Neural Computing and Applications, 2024

   Abs arXiv Bib Code

   The broad usage of mobile devices nowadays, the sensitiveness of the information contained in them, and the shortcomings of current mobile user authentication methods are calling for novel, secure, and unobtrusive solutions to verify the users’ identity. In this article, we propose TypeFormer, a novel transformer architecture to model free-text keystroke dynamics performed on mobile devices for the purpose of user authentication. The proposed model consists in temporal and channel modules enclosing two long short-term memory recurrent layers, Gaussian range encoding, a multi-head self-attention mechanism, and a block-recurrent transformer layer. Experimenting on one of the largest public databases to date, the Aalto mobile keystroke database, TypeFormer outperforms current state-of-the-art systems achieving equal error rate values of 3.25% using only five enrolment sessions of 50 keystrokes each. In such way, we contribute to reducing the traditional performance gap of the challenging mobile free-text scenario with respect to its desktop and fixed-text counterparts. To highlight the design rationale, an analysis of the experimental results of the different modules implemented in the development of TypeFormer is carried out. Additionally, we analyse the behaviour of the model with different experimental configurations such as the length of the keystroke sequences and the amount of enrolment sessions, showing margin for improvement.

   @article{stragapede2024typeformer,
     title = {TypeFormer: Transformers for mobile keystroke biometrics},
     author = {Stragapede, Giuseppe and Delgado-Santos, Paula and Tolosana, Ruben and Vera-Rodriguez, Ruben and Guest, Richard and Morales, Aythami},
     journal = {Neural Computing and Applications},
     volume = {36},
     number = {29},
     pages = {18531--18545},
     year = {2024},
     publisher = {Springer},
   }

2. Biometric recognition based on mobile human-computer interaction

   Giuseppe Stragapede

   Universidad Autónoma de Madrid, 2024

   Abs DOI Bib

   The rapid digitalization of society is creating unprecedented Human-Computer Interaction (HCI) scenarios. Mobile devices such as smartphones and wearables have high computing and connectivity capabilities, and they are provided with several sensors that are able to acquire a vast and diverse amount of information pertaining to the users, for purposes such as security (biometric verification systems), health and fitness (activity trackers), user profiling (social media and e-commerce), among others. From one perspective, many studies have highlighted the disadvantages of passwords and physiological biometric characteristic such as fingerprint or face, as they may be easy to be stolen, forged, and they cannot guarantee prolonged protection throughout the entire device usage. To this end, it can be shown that the touch gestures on the smartphone screen and the body movements captured by the background sensors are traits that provide enough discriminative power to be associated with users’ identities, and therefore to be used for biometric recognition. On the flip side, the large availability of personal data generated on mobile devices has turned this technology into a potential source of major invasion of personal privacy. In fact, thanks to the recent advancements of Artificial Intelligence (AI), the automated processing of mobile user interaction data can easily reveal users’ sensitive attributes, reducing the privacy and security of the users. In this scenario, this Thesis work aims at advancing behavioral biometrics for mobile transparent user authentication. This objective is pursued by considering different modalities (touch gestures, mobile sensor data patterns, and keystroke dynamics), state-of-the-art deep learning classifiers, metrics, and databases. At the same time, individuals’ mobile behavioral biometric data used for authentication might enclose personal and sensitive information. This aspect is also explored for privacy quantification. This Dissertation comprises four different parts. Part I first concentrates on the problem statement and main contributions of the Thesis. The experimental chapters are then divided into two parts, Part II, and Part III. Lastly, Part IV concludes the Thesis. Part I presents the basics of biometric systems, together with an explanation of the theoretical framework and the practical applications of the current Thesis, an outline of the Dissertation, and a summary of the research contributions originated from this Thesis. Then, the most important aspects of related work are described, with a presentation of the databases used, and the metrics adopted in the experimental part of this Dissertation. The first experimental part (Part II) presents a comparative analysis of unimodal and multimodal behavioral biometric traits acquired while the subjects perform different activities on the phone such as typing, scrolling, drawing a number, and tapping on the screen, considering the touchscreen and the simultaneous background sensor data (accelerometer, gravity sensor, gyroscope, linear accelerometer, and magnetometer). The experiments are performed over HuMIdb, one of the largest and most comprehensive freely available mobile user interaction databases to date. A separate Recurrent Neural Network (RNN) with triplet loss is implemented for each single modality, followed by the biometric fusion at score level, leading to Equal Error Rates (EER) ranging from 4% to 9% depending on the modality combination in a 3-second interval. Then, a new database, BehavePassDB, collected within this Thesis work, is presented and benchmarked with similar results in terms of recognition performance. BehavePassDB is structured into separate acquisition sessions and tasks to mimic the most common aspects of mobile Human-Computer Interaction (HCI), and it was acquired through a dedicated mobile app installed on the subjects devices, also including the case of different users on the same device for evaluation. An international ongoing competition, MobileB2C, was organized based on the novel database. In the second experimental part (Part III of this Dissertation), the focus of exploring mobile behavioral biometrics is narrowed down to Keystroke Dynamics (KD), which resulted to be the most discriminative trait among the ones considered in the earlier chapters. First, a novel Transformer architecture, TypeFormer, is proposed for mobile KD-based verification improving recent state-of-the-art keystroke verification systems based on LSTM RNNs. Then, a novel experimental framework to benchmark keystroke for biometric verification is described, designed to quantify the recognition performance as well as the fairness of biometric systems. The framework is provided in the form of the Keystroke Verification Challenge at the 2023 IEEE International Conference on Big Data. To create the framework, we consider two of the largest public databases of keystroke dynamics up to date, the Aalto Desktop and Mobile Keystroke Databases, extracting datasets that guarantee a minimum amount of data per subject, age and gender annotations, absence of corrupted data, and that avoid too unbalanced subject distributions with respect to the considered demographic attributes. The framework is designed to represent the modern challenges of massive application usage, counting on over 185,000 subjects, and it considers tweet-long sequences of arbitrary text, in mobile and desktop scenarios.

   @phdthesis{stragapede2024biometric,
     title = {Biometric recognition based on mobile human-computer interaction},
     author = {Stragapede, Giuseppe},
     year = {2024},
     school = {Universidad Aut{\'o}noma de Madrid},
     doi = {http://hdl.handle.net/10486/715062},
   }

2023

1. IEEE BigData 2023 Keystroke Verification Challenge (KVC)

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 27 more authors

   In 2023 IEEE International Conference on Big Data (BigData), 2023

   Abs DOI arXiv Bib

   This paper describes the results of the IEEE BigData 2023 Keystroke Verification Challenge (KVC), that considers the biometric verification performance of Keystroke Dynamics (KD), captured as tweet-long sequences of variable transcript text from over 185,000 subjects. The data are obtained from two of the largest public databases of KD up to date, the Aalto Desktop and Mobile Keystroke Databases, guaranteeing a minimum amount of data per subject, age and gender annotations, absence of corrupted data, and avoiding excessively unbalanced subject distributions with respect to the considered demographic attributes. Several neural architectures were proposed by the participants, leading to global Equal Error Rates (EERs) as low as 3.33% and 3.61% achieved by the best team respectively in the desktop and mobile scenario, outperforming the current state of the art biometric verification performance for KD. Hosted on CodaLab, the KVC will be made ongoing to represent a useful tool for the research community to compare different approaches under the same experimental conditions and to deepen the knowledge of the field.

   @inproceedings{stragapede2023ieee,
     title = {IEEE BigData 2023 Keystroke Verification Challenge (KVC)},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and DeAndres-Tame, Ivan and Damer, Naser and Fierrez, Julian and Ortega-Garcia, Javier and Acien, Alejandro and Gonzalez, Nahuel and Shadrikov, Andrei and Gordin, Dmitrii and Schmitt, Leon and Wimmer, Daniel and Großmann, Christoph and Krieger, Joerdis and Heinz, Florian and Krestel, Ron and Mayer, Christoffer and Haberl, Simon and Gschrey, Helena and Yamagishi, Yosuke and Saha, Sanjay and Rasnayaka, Sanka and Wickramanayake, Sandareka and Sim, Terence and Gutfeter, Weronika and Baran, Adam and Krzysztoń, Mateusz and Jaskóła, Przemysław},
     booktitle = {2023 IEEE International Conference on Big Data (BigData)},
     pages = {6092--6100},
     year = {2023},
     organization = {IEEE},
     doi = {https://doi.org/10.1109/BigData59044.2023.10386557},
   }

2. Keystroke Verification Challenge (KVC): Biometric and Fairness Benchmark Evaluation

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 4 more authors

   IEEE Access, 2023

   Abs DOI arXiv Bib Supp Code Website

   Analyzing keystroke dynamics (KD) for biometric verification has several advantages: it is among the most discriminative behavioral traits; keyboards are among the most common human-computer interfaces, being the primary means for users to enter textual data; its acquisition does not require additional hardware, and its processing is relatively lightweight; and it allows for transparently recognizing subjects. However, the heterogeneity of experimental protocols and metrics, and the limited size of the databases adopted in the literature impede direct comparisons between different systems, thus representing an obstacle in the advancement of keystroke biometrics. To alleviate this aspect, we present a new experimental framework to benchmark KD-based biometric verification performance and fairness based on tweet -long sequences of variable transcript text from over 185,000 subjects, acquired through desktop and mobile keyboards, extracted from the Aalto Keystroke Databases. The framework runs on CodaLab in the form of the Keystroke Verification Challenge (KVC). Moreover, we also introduce a novel fairness metric, the Skewed Impostor Ratio (SIR), to capture inter - and intra -demographic group bias patterns in the verification scores. We demonstrate the usefulness of the proposed framework by employing two state-of-the-art keystroke verification systems, TypeNet and TypeFormer, to compare different sets of input features, achieving a less privacy-invasive system, by discarding the analysis of text content (ASCII codes of the keys pressed) in favor of extended features in the time domain. Our experiments show that this approach allows to maintain satisfactory performance.

   @article{ieeeaccess,
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and Damer, Naser and Fierrez, Julian and Ortega-Garcia, Javier},
     journal = {IEEE Access},
     title = {Keystroke Verification Challenge (KVC): Biometric and Fairness Benchmark Evaluation},
     year = {2023},
     volume = {12},
     number = {},
     pages = {1102-1116},
     doi = {10.1109/ACCESS.2023.3345452},
   }

3. Workshop on Advances of Mobile and Wearable Biometrics (WAMWB)

   Giuseppe Stragapede, Ruben Vera-Rodriguez, and Ruben Tolosana

   In Proceedings of the 25th International Conference on Mobile Human-Computer Interaction, 2023

   Abs DOI Bib PDF Supp Website

   Biometrics is defined as the automated recognition of individuals based on their biological and behavioural characteristics. It represents a fundamental aspect of mobile Human-Computer Interaction (HCI), as mobile devices such as smartphones and wearables are designed to capture, process and transmit biometric data. While people benefit from the innumerable applications of biometric data in the context of HCI, new concerns have raised in relation with the performance, reliability, protection of privacy, bias, misuse, regulations, and their impact on society. The Workshop on Advances of Mobile and Wearable Biometrics (WAMWB) aims to highlight recent developments with respect to such challenges and risks.

   @inproceedings{stragapede2023workshop,
     title = {Workshop on Advances of Mobile and Wearable Biometrics (WAMWB)},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben},
     booktitle = {Proceedings of the 25th International Conference on Mobile Human-Computer Interaction},
     pages = {1--3},
     year = {2023},
     doi = {https://doi.org/10.1145/3565066.3609792},
   }

4. BehavePassDB: public database for mobile behavioral biometrics and benchmark evaluation

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 1 more author

   Pattern Recognition, 2023

   Abs DOI arXiv Bib Video Code

   Mobile behavioral biometrics have become a popular topic of research, reaching promising results in terms of authentication, exploiting a multimodal combination of touchscreen and background sensor data. However, there is no way of knowing whether state-of-the-art classifiers in the literature can distinguish between the notion of user and device. In this article, we present a new database, BehavePassDB, structured into separate acquisition sessions and tasks to mimic the most common aspects of mobile Human-Computer Interaction (HCI). BehavePassDB is acquired through a dedicated mobile app installed on the subjects devices, also including the case of different users on the same device for evaluation. We propose a standard experimental protocol and benchmark for the research community to perform a fair comparison of novel approaches with the state of the art1. We propose and evaluate a system based on Long-Short Term Memory (LSTM) architecture with triplet loss and modality fusion at score level.

   @article{stragapede2023behavepassdb,
     title = {BehavePassDB: public database for mobile behavioral biometrics and benchmark evaluation},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami},
     journal = {Pattern Recognition},
     volume = {134},
     pages = {109089},
     year = {2023},
     publisher = {Elsevier},
     doi = {https://doi.org/10.1016/j.patcog.2022.109089},
   }

5. Mobile keystroke biometrics using transformers

   Giuseppe Stragapede, Paula Delgado-Santos, Ruben Tolosana, and 3 more authors

   In 2023 IEEE 17th International Conference on Automatic Face and Gesture Recognition (FG), 2023

   Abs DOI arXiv Bib

   Among user authentication methods, behavioural biometrics has proven to be effective against identity theft as well as user-friendly and unobtrusive. One of the most popular traits in the literature is keystroke dynamics due to the large deployment of computers and mobile devices in our society. This paper focuses on improving keystroke biometric systems on the free-text scenario. This scenario is characterised as very challenging due to the uncontrolled text conditions, the influence of the user’s emotional and physical state, and the in-use application. To overcome these drawbacks, methods based on deep learning such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) have been proposed in the literature, outperforming traditional machine learning methods. However, these architectures still have aspects that need to be reviewed and improved. To the best of our knowledge, this is the first study that proposes keystroke biometric systems based on Transformers. The proposed Transformer architecture has achieved Equal Error Rate (EER) values of 3.84% in the popular Aalto mobile keystroke database using only 5 enrolment sessions, outperforming by a large margin other state-of-the-art approaches in the literature.

   @inproceedings{stragapede2023mobile,
     title = {Mobile keystroke biometrics using transformers},
     author = {Stragapede, Giuseppe and Delgado-Santos, Paula and Tolosana, Ruben and Vera-Rodriguez, Ruben and Guest, Richard and Morales, Aythami},
     booktitle = {2023 IEEE 17th International Conference on Automatic Face and Gesture Recognition (FG)},
     pages = {1--6},
     year = {2023},
     organization = {IEEE},
     doi = {https://doi.org/10.1109/FG57933.2023.10042710}
   }

2022

1. A survey of privacy vulnerabilities of mobile device sensors

   Paula Delgado-Santos, Giuseppe Stragapede, Ruben Tolosana, and 3 more authors

   ACM Computing Surveys (CSUR), 2022

   Abs DOI arXiv Bib

   The number of mobile devices, such as smartphones and smartwatches, is relentlessly increasing, to almost 6.8 billion by 2022, and along with it, the amount of personal and sensitive data captured by them. This survey overviews the state of the art of what personal and sensitive user attributes can be extracted from mobile device sensors, emphasizing critical aspects such as demographics, health and body features, activity and behavior recognition, and so forth. In addition, we review popular metrics in the literature to quantify the degree of privacy and discuss powerful privacy methods to protect the sensitive data while preserving data utility for analysis. Finally, open research questions are presented for further advancements in the field.

   @article{delgado2022survey,
     title = {A survey of privacy vulnerabilities of mobile device sensors},
     author = {Delgado-Santos, Paula and Stragapede, Giuseppe and Tolosana, Ruben and Guest, Richard and Deravi, Farzin and Vera-Rodriguez, Ruben},
     journal = {ACM Computing Surveys (CSUR)},
     volume = {54},
     number = {11s},
     pages = {1--30},
     year = {2022},
     publisher = {ACM New York, NY},
     doi = {https://doi.org/10.1145/3510579}
   }

2. Mobile behavioral biometrics for passive authentication

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 3 more authors

   Pattern Recognition Letters, 2022

   Abs DOI arXiv Bib

   Current mobile user authentication systems based on PIN codes, fingerprint, and face recognition have several shortcomings. Such limitations have been addressed in the literature by exploring the feasibility of passive authentication on mobile devices through behavioral biometrics. In this line of research, this work carries out a comparative analysis of unimodal and multimodal behavioral biometric traits acquired while the subjects perform different activities on the phone such as typing, scrolling, drawing a number, and tapping on the screen, considering the touchscreen and the simultaneous background sensor data (accelerometer, gravity sensor, gyroscope, linear accelerometer, and magnetometer). Our experiments are performed over HuMIdb,1 one of the largest and most comprehensive freely available mobile user interaction databases to date. A separate Recurrent Neural Network (RNN) with triplet loss is implemented for each single modality. Then, the weighted fusion of the different modalities is carried out at score level. In our experiments, the most discriminative background sensor is the magnetometer, whereas among touch tasks the best results are achieved with keystroke in a fixed-text scenario. In all cases, the fusion of modalities is very beneficial, leading to Equal Error Rates (EER) ranging from 4% to 9% depending on the modality combination in a 3-second interval.

   @article{stragapede2022mobile,
     title = {Mobile behavioral biometrics for passive authentication},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and Acien, Alejandro and Le Lan, Ga{\"e}l},
     journal = {Pattern Recognition Letters},
     volume = {157},
     pages = {35--41},
     year = {2022},
     publisher = {Elsevier},
     doi = {https://doi.org/10.1016/j.patrec.2022.03.014}
   }

3. IJCB 2022 mobile behavioral biometrics competition (MobileB2C)

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 8 more authors

   In 2022 IEEE International Joint Conference on Biometrics (IJCB), 2022

   Abs DOI arXiv Bib Supp Website

   This paper describes the experimental framework and results of the IJCB 2022 Mobile Behavioral Biometrics Competition (MobileB2C). The aim of MobileB2C is benchmarking mobile user authentication systems based on behavioral biometric traits transparently acquired by mobile devices during ordinary Human-Computer Interaction (HCI), using a novel public database, BehavePassDB, and a standard experimental protocol. The competition is divided into four tasks corresponding to typical user activities: keystroke, text reading, gallery swiping, and tapping. The data are composed of touchscreen data and several background sensor data simultaneously acquired. "Random" (different users with different devices) and "skilled" (different user on the same device attempting to imitate the legitimate one) impostor scenarios are considered. The results achieved by the participants show the feasibility of user authentication through behavioral biometrics, although this proves to be a non-trivial challenge. MobileB2C will be established as an on-going competition.

   @inproceedings{stragapede2022ijcb,
     title = {IJCB 2022 mobile behavioral biometrics competition (MobileB2C)},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and Fierrez, Julian and Ortega-Garcia, Javier and Rasnayaka, Sanka and Seneviratne, Sachith and Dissanayake, Vipula and Liebers, Jonathan and others},
     booktitle = {2022 IEEE International Joint Conference on Biometrics (IJCB)},
     pages = {1--7},
     year = {2022},
     organization = {IEEE},
     doi = {https://doi.org/10.1109/IJCB54206.2022.10007985},
   }

4. Mobile passive authentication through touchscreen and background sensor data

   Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, and 3 more authors

   In 2022 International Workshop on Biometrics and Forensics (IWBF), 2022

   Abs DOI Bib PDF

   The security and usability shortcomings of current mobile user authentication systems based on PIN codes, fingerprint, and face recognition are well known. To overcome such limitations, the present work focuses on the comparative analysis of unimodal and multimodal behavioral biometric traits suitable for mobile passive authentication, such as touchscreen data during separate gestures (keystroke, scrolling, drawing a number, tapping on the screen), and background sensor data (accelerometer, gravity sensor, gyroscope, linear accelerometer, magnetometer).This paper carries out a performance evaluation over one of the most complete and challenging databases to date with mobile user interaction data, HuMIdb, with 600 subjects. For each individual modality, we propose a separate RNN (Recurrent Neural Network) trained with semi-hard triplet loss. In addition, we perform the fusion of the different modalities at score level. Our results show that the best performing tasks are keystroke and drawing a number, whereas the most discriminative background sensor is the magnetometer. Additionally, the fusion of modalities is very beneficial, consistently reducing the Equal Error Rates (EER) by half (ranging from 5% to 13% depending on the modality combination).

   @inproceedings{stragapede2022mobilf,
     title = {Mobile passive authentication through touchscreen and background sensor data},
     author = {Stragapede, Giuseppe and Vera-Rodriguez, Ruben and Tolosana, Ruben and Morales, Aythami and Acien, Alejandro and Le Lan, Ga{\"e}l},
     booktitle = {2022 International Workshop on Biometrics and Forensics (IWBF)},
     pages = {1--6},
     year = {2022},
     organization = {IEEE},
     doi = {https://doi.org/10.1109/IWBF55382.2022.9794524}
   }