FACES : facial analysis with compressed efficient systems

Lajić, Romanela (author), Peer, Peter (author), Štruc, Vitomir (author), Han, Dong Seog (author), Meden, Blaž (author), Emeršič, Žiga (author)

Due to their promising performance vision transformers are increasingly being incorporated into various biometric solutions, mainly in the domain of face analysis. However, their size and computational expense remain the biggest challenge when it comes to their full utilization and there is a high demand for optimization of these models. In this paper we propose a novel pruning technique for face analysis vision transformers aimed at reducing their memory and computational cost. The method uses existing transformer parameters as importance scores, which allows for a simple one-shot pruning and retraining approach. By testing the method on the SWINFace transformer for both verification and attribute recognition tasks, we show that the models compressed up to 50% sparsity level maintain the performance or even outperform the original model, while also outperforming state-of-the-art vision transformer pruning methods and showing versatility for different face analysis tasks.

• Language: English
• Keywords: face analysis, neural network compression, pruning, vision transformers
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: FRI - Faculty of Computer and Information Science
• Publication status: Published
• Publication version: Version of Record
• Year: 2026
• Number of pages: 5 str.
• Numbering: Vol. , no.
• PID: 20.500.12556/RUL-181219
• UDC: 004.93:57.087.1
• ISSN on article: 2405-9595
• DOI: 10.1016/j.icte.2026.02.008
• COBISS.SI-ID: 273178115
• Publication date in RUL: 27.03.2026

Record is a part of a journal

• Title: ICT express
• Publisher: Elsevier
• ISSN: 2405-9595
• COBISS.SI-ID: 526132505

Licences

• License: CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
• Link: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Description: The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.

Secondary language

• Language: Slovenian
• Keywords: analiza obraza, kompresija nevronskih mrež, obrezovanje, vidni transformatorji, ViT

Projects

• Funder: ARIS - Slovenian Research and Innovation Agency
• Project number: P2-0250-2018
• Name: Metrologija in biometrični sistemi

• Funder: EC - European Commission
• Project number: 101225635
• Name: On-the-move Schengen border control using extended EUDI wallet, smartphone and external sensor technologies
• Acronym: OnMoveID