Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition
The EU AI Act underscores the importance of transparency, user-centricity, and robustness in AI systems, particularly for high-risk applications. In response, we present advancements in XentricAI, an explainable hand gesture recognition (HGR) system designed to meet these regulatory requirements. Xe...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-06-01
|
| Series: | Machine Learning with Applications |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666827025000386 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849691487312281600 |
|---|---|
| author | Sarah Seifi Tobias Sukianto Cecilia Carbonelli Lorenzo Servadei Robert Wille |
| author_facet | Sarah Seifi Tobias Sukianto Cecilia Carbonelli Lorenzo Servadei Robert Wille |
| author_sort | Sarah Seifi |
| collection | DOAJ |
| description | The EU AI Act underscores the importance of transparency, user-centricity, and robustness in AI systems, particularly for high-risk applications. In response, we present advancements in XentricAI, an explainable hand gesture recognition (HGR) system designed to meet these regulatory requirements. XentricAI addresses fundamental challenges in HGR, such as the opacity of black-box models using explainable AI methods and the handling of distributional shifts in real-world data through transfer learning techniques.We extend an existing radar-based HGR dataset by adding 28,000 new gestures, with contributions from multiple users across varied locations, including 24,000 out-of-distribution gestures. Leveraging this real-world dataset, we enhance XentricAI’s capabilities by integrating a variational autoencoder module for improved gesture anomaly detection, incorporating user-specific dynamic thresholding. This integration enables the identification of 11.50% more anomalous gestures.Our extensive evaluations demonstrate a 97.5% success rate in characterizing these anomalies, significantly improving system explainability. Furthermore, the implementation of transfer learning techniques has shown a substantial increase in user adaptability, with an average performance improvement of at least 15.17%.This work contributes to the development of trustworthy AI systems by providing both technical advancements and regulatory compliance, offering a commercially viable solution that aligns with the EU AI Act requirements. |
| format | Article |
| id | doaj-art-c56838930ad543ccb6d182e0b81bfb7f |
| institution | DOAJ |
| issn | 2666-8270 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Machine Learning with Applications |
| spelling | doaj-art-c56838930ad543ccb6d182e0b81bfb7f2025-08-20T03:21:01ZengElsevierMachine Learning with Applications2666-82702025-06-012010065510.1016/j.mlwa.2025.100655Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognitionSarah Seifi0Tobias Sukianto1Cecilia Carbonelli2Lorenzo Servadei3Robert Wille4Chair for Design Automation, Technical University Munich, Arcisstr.21, Munich, 80333, Bavaria, Germany; Infineon Technologies AG, Am Campeon 1-15, Neubiberg, 80939, Bavaria, Germany; Corresponding author at: Chair for Design Automation, Technical University Munich, Arcisstr.21, Munich, 80333, Bavaria, Germany.Infineon Technologies AG, Am Campeon 1-15, Neubiberg, 80939, Bavaria, Germany; Institute for Signal Processing, Johannes Kepler University Linz, Altenbergerstraße 69, Linz, 4040, AustriaInfineon Technologies AG, Am Campeon 1-15, Neubiberg, 80939, Bavaria, GermanyChair for Design Automation, Technical University Munich, Arcisstr.21, Munich, 80333, Bavaria, GermanyChair for Design Automation, Technical University Munich, Arcisstr.21, Munich, 80333, Bavaria, Germany; Software Competence Center Hagenberg GmbH (SCCH), Softwarepark 32a, Hagenberg, 4232, AustriaThe EU AI Act underscores the importance of transparency, user-centricity, and robustness in AI systems, particularly for high-risk applications. In response, we present advancements in XentricAI, an explainable hand gesture recognition (HGR) system designed to meet these regulatory requirements. XentricAI addresses fundamental challenges in HGR, such as the opacity of black-box models using explainable AI methods and the handling of distributional shifts in real-world data through transfer learning techniques.We extend an existing radar-based HGR dataset by adding 28,000 new gestures, with contributions from multiple users across varied locations, including 24,000 out-of-distribution gestures. Leveraging this real-world dataset, we enhance XentricAI’s capabilities by integrating a variational autoencoder module for improved gesture anomaly detection, incorporating user-specific dynamic thresholding. This integration enables the identification of 11.50% more anomalous gestures.Our extensive evaluations demonstrate a 97.5% success rate in characterizing these anomalies, significantly improving system explainability. Furthermore, the implementation of transfer learning techniques has shown a substantial increase in user adaptability, with an average performance improvement of at least 15.17%.This work contributes to the development of trustworthy AI systems by providing both technical advancements and regulatory compliance, offering a commercially viable solution that aligns with the EU AI Act requirements.http://www.sciencedirect.com/science/article/pii/S2666827025000386Machine learning (ML)Explainable AI (XAI)Hand gesture recognition (HGR)Frequency-modulated continuous wave (FMCW) radar |
| spellingShingle | Sarah Seifi Tobias Sukianto Cecilia Carbonelli Lorenzo Servadei Robert Wille Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition Machine Learning with Applications Machine learning (ML) Explainable AI (XAI) Hand gesture recognition (HGR) Frequency-modulated continuous wave (FMCW) radar |
| title | Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition |
| title_full | Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition |
| title_fullStr | Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition |
| title_full_unstemmed | Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition |
| title_short | Complying with the EU AI Act: Innovations in explainable and user-centric hand gesture recognition |
| title_sort | complying with the eu ai act innovations in explainable and user centric hand gesture recognition |
| topic | Machine learning (ML) Explainable AI (XAI) Hand gesture recognition (HGR) Frequency-modulated continuous wave (FMCW) radar |
| url | http://www.sciencedirect.com/science/article/pii/S2666827025000386 |
| work_keys_str_mv | AT sarahseifi complyingwiththeeuaiactinnovationsinexplainableandusercentrichandgesturerecognition AT tobiassukianto complyingwiththeeuaiactinnovationsinexplainableandusercentrichandgesturerecognition AT ceciliacarbonelli complyingwiththeeuaiactinnovationsinexplainableandusercentrichandgesturerecognition AT lorenzoservadei complyingwiththeeuaiactinnovationsinexplainableandusercentrichandgesturerecognition AT robertwille complyingwiththeeuaiactinnovationsinexplainableandusercentrichandgesturerecognition |