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

arXiv: https://arxiv.org/abs/2405.01088 · HTML: https://mlsanity.com/paper/2405.01088/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 69 references that are in the map.
- First posted on arXiv 2 May 2024.

## Builds on (most cited first, 17 of 17)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [How Good is ChatGPT at Face Biometrics? A First Look into Recognition, Soft Biometrics, and Explainability](https://mlsanity.com/paper/2401.13641/) (2024)
- [DeepWriteSYN: On-Line Handwriting Synthesis via Deep Short-Term Representations](https://mlsanity.com/paper/2009.06308/) (2020)
- [BiosecurID: a multimodal biometric database](https://mlsanity.com/paper/2111.03472/) (2021)
- [BioTouchPass2: Touchscreen Password Biometrics Using Time-Aligned Recurrent Neural Networks](https://mlsanity.com/paper/2001.10223/) (2020)
- [TypeNet: Deep Learning Keystroke Biometrics](https://mlsanity.com/paper/2101.05570/) (2021)
- [SVC-onGoing: Signature Verification Competition](https://mlsanity.com/paper/2108.06090/) (2021)
- [SetMargin Loss applied to Deep Keystroke Biometrics with Circle Packing Interpretation](https://mlsanity.com/paper/2109.00938/) (2021)
- [Synthetic Data for the Mitigation of Demographic Biases in Face Recognition](https://mlsanity.com/paper/2402.01472/) (2024)
- [BehavePassDB: Public Database for Mobile Behavioral Biometrics and Benchmark Evaluation](https://mlsanity.com/paper/2206.02502/) (2022)
- [TypeFormer: Transformers for Mobile Keystroke Biometrics](https://mlsanity.com/paper/2212.13075/) (2022)
- [Free-Text Keystroke Dynamics for User Authentication](https://mlsanity.com/paper/2107.07009/) (2021)
- [Keystroke Verification Challenge (KVC): Biometric and Fairness Benchmark Evaluation](https://mlsanity.com/paper/2311.06000/) (2023)
- [Keystroke Biometrics in Response to Fake News Propagation in a Global Pandemic](https://mlsanity.com/paper/2005.07688/) (2020)
- [BeCAPTCHA-Type: Biometric Keystroke Data Generation for Improved Bot Detection](https://mlsanity.com/paper/2207.13394/) (2022)
- [IEEE BigData 2023 Keystroke Verification Challenge (KVC)](https://mlsanity.com/paper/2401.16559/) (2024)

## Built on by (most cited first, 5 of 5)

- [KVC-onGoing: Keystroke Verification Challenge](https://mlsanity.com/paper/2412.20530/) (2024)
- [Is It Really You? Exploring Biometric Verification Scenarios in Photorealistic Talking-Head Avatar Videos](https://mlsanity.com/paper/2508.00748/) (2025)
- [Leveraging Avatar Fingerprinting: A Multi-Generator Photorealistic Talking-Head Public Database and Benchmark](https://mlsanity.com/paper/2603.26934/) (2026)
- [Exploring Vision-Language Models for Online Signature Verification: A Zero-Shot Capability Study](https://mlsanity.com/paper/2605.14845/) (2026)
- [EERLoss: A Novel Loss Function for Training Deep Biometric Models. A Case Study in Keystroke Dynamics](https://mlsanity.com/paper/2606.24586/) (2026)

## Abstract (arXiv)

In 2021, the pioneering work 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 typing 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 Equal Error Rates (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 fixed global threshold for all subjects, the EERs are respectively 3.25% and 3.61% for desktop and mobile scenarios, outperforming previous approaches by a significant margin. The source code for dataset generation, model, and training process is publicly available.