Results and Achievements

2024-07-16

Horizon Europe project approved – ACHILLES

Authors: VisTeam

We are thrilled to announce that ACHILLES Project, a project where VisTeam (ISR-Coimbra) participates in, was approved to be financed by Horizon Europe, in a total budget for ISR-Coimbra of 470k€ (total eligible budget of the project is approximately 9.6M€). The ISR-Coimbra team will be leaded by Prof. Nuno Gonçalves.
ISR-Coimbra participation is related to the building of privacy-by-design learning models and a use case related to Biometrics.
The project has its main focus on building Deep Learning models that are more explainable, fair, demographically balanced and with data whose sources are perfectly identified, with more security and privacy. This project aims to build the foundations for auditability of Deep Learning algorithms, one of the focuses of the European Commission regarding the AI ​​Act.
The consortium is led by Fraunhofer Portugal and has a total of 16 participants.

A brief description about ACHILLES Human-Centred Machine Learning: Lighter, Clearer, Safer:
Inspired by the Olympic motto “Citius, Altius, Fortius” (Faster, Higher, Stronger),
ACHILLES delivers a modular framework for developing ML-based systems that are Lighter, Clearer, and Safer, inherently aligned with a broad spectrum of compliance requirements:

Lighter: ACHILLES strives for sustainable AI systems by promoting less data-intensive and energy consuming models without compromising output quality, in alignment with the European Green Deal.

Clearer: ACHILLES underscores transparency and explainability, critical for ethical compliance, aligning with the European Approach to AI’s focus on ethical principles, accountability, and trustworthiness.

Safer: ACHILLES commits to improving robustness, safety, and reliability in ML applications, ensuring the legal and technical compliance necessary for the safe integration of AI solutions across various sectors.

ACHILLES adopts an iterative cycle inspired by the phased approach of clinical trials, from preclinical explorations to post-market surveillance. This cycle addresses different stages (as seen in Figure 1):

Human-centric (start): prioritizing the human perspective, this stage employs value-sensitive design to involve users from the outset, breaking ground for AI that aligns with human values and regulatory expectations.

Data-centric: this stage tackles data quality, bias, synthetic data generation, and interpretable concepts, ensuring compliance with regulations like GDPR and addressing privacy and security concerns.

Model-centric: here, ACHILLES aims to improve training efficiency and privacy-preserving methodologies, such as federated learning.

Deployment-centric: this stage emphasizes energy efficiency in the deployment stage, offering techniques to minimize energy consumption and optimise models for existing infrastructure.

Human-centric (end): the cycle recentres on the human perspective, underscoring human-AI interaction and agency. This encompasses explainability, continuous monitoring, and transparent reporting through augmented data and model cards, akin to drug leaflets in the clinical trials analogy.

We also propose an ACHILLES Integrated Development Environment (IDE) that aims to be a new standard for machine learning development by providing a comprehensive framework empowering scientists, developers, and engineers to embed the project’s vision into AI products seamlessly, making the right thing the easiest thing to do. By incorporating these principles, ACHILLES contributes to the three pillars of a robust compliance framework by: a) studying and advancing current compliance targets in legal, ethical, and sustainability domains; b) providing technical modules and a specialized IDE to promote best practices and seamless adherence to targets; and c) methods to verify compliance, through exhaustive and transparent reporting. ACHILLES seeks to help creating efficient, transparent, and trustworthy-by-design AI solutions that empower human-centred decision-making while addressing the complex challenges of AI’s societal impact with a focus on sustainability and ethical integrity.

Related Content

Researcher Coordinator, VIS TEAM Leader
Post-Doc Researcher and Project Manager
No tagged content to show
No tagged content to show
No tagged content to show

sugested news

Laser engraving of precious metal artifacts (UniqueMark® deterministic...
UniqueMark® and UniQode® Glitter patent published
Paper about protecting facial recognition systems against morphing...

RECENT PROJECTS

FACING2 – Face Image Understanding
VISUAL-ID – Unique Visual Identities in Graphics, Images and Faces
UniqueMark

RECENT PUBLICATIONS

MorFacing: A Benchmark for Estimation Face Recognition Robustness to Face Morphing Attacks

Authors: Iurii Medvedev and Nuno Gonçalves
Featured in: IEEE International Joint Conference on Biometrics (IJCB 2024)

Face Liveness Detection Competition (LivDet-Face)

Authors: Lambert Igene, Afzal Hossain, Stephanie Schuckers, Mohammad Zahir Uddin Chowdhury, Humaira Rezaie, Ayden Rollins, Jesse Dykes, Rahul Vijaykumar, Sebastien Marcel, Juan Tapia, Carlos Aravena, Daniel Schulz, Nima Karimian and Anafsheh Adami, Diogo Nunes, João Marcos, Nuno Gonçalves, Lovro Sikošek, Borut Batagelj, Nima Schei, David Pabon, Manuela Tiedemann, Vasiliy Pryadchenko, Aleksandr Alenin, Alhasan Alkhaddour, Anton Pimenov, Artem Tregubov, Igor Avdonin, Maxim Lazantsev and Mikhail Pozigun
Featured in: IEEE International Joint Conference on Biometrics Competitions, 2024

Social NSTransformers: Low-Quality Pedestrian Trajectory Prediction

Authors: Zihan Jiang, Yiqun Ma, Bingyu Shi, Xin Lu, Jian Xing, Nuno Gonçalves and Bo Jin
Featured in: IEEE Transactions on Artificial Intelligence

Institute of Systems and Robotics Department of Electrical and Computers Engineering University of Coimbra