{"id":36921,"date":"2026-09-30T15:38:30","date_gmt":"2026-09-30T13:38:30","guid":{"rendered":"https:\/\/www.iti.gr\/iti\/?page_id=36921"},"modified":"2026-09-30T15:42:39","modified_gmt":"2026-09-30T13:42:39","slug":"best-publications-awards","status":"publish","type":"page","link":"https:\/\/www.iti.gr\/iti\/en\/best-publications-awards\/","title":{"rendered":"Best Publications Awards"},"content":{"rendered":"<div class=\"container mt-5\">\n<div class=\"award-layout\">\n<nav class=\"award-toc\" aria-label=\"\u03a0\u03b5\u03c1\u03b9\u03b5\u03c7\u03cc\u03bc\u03b5\u03bd\u03b1 \u03c3\u03b5\u03bb\u03af\u03b4\u03b1\u03c2\">\n<ul>\n<li><a href=\"#awards-2024\">Best Publications 2024<\/a><\/li>\n<li><a href=\"#awards-2023\">Best Publications 2023<\/a><\/li>\n<li><a href=\"#award-process\">Award Selection Process<\/a><\/li>\n<\/ul>\n<\/nav>\n<div class=\"award-main\">\n<div class=\"container\">\n<div class=\"row\">\n<div class=\"col-lg-12 col-md-12 col-12 position-title\">\n<h2 id=\"awards-2024\" class=\"award-year\"><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> <strong>Best Publications 2024<\/strong> <i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i><\/h2>\n<\/div>\n<\/div>\n<h5><em><strong>Best Journal Publication Award<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Apostolos Evangelidis \u2013 Efficient deep Q-learning for industrial equipment calibration in elevator manufacturing<\/b><span class=\"award-citation\">A. Evangelidis, N. Dimitriou, P. Charalampous, T. D. Mastos, D. Tzovaras, &#8220;<i>Efficient deep Q-learning for industrial equipment calibration in elevator manufacturing.<\/i>&#8220;, IEEE Transactions on Industrial Informatics, vol. 20, no. 10, pp. 12220 &#8211; 12230, October 2024.<\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 Excellent originality, devising novel mathematical formulation and subsequently applying to real manufacturing facilities for extensive testing.<\/p>\n<p><strong>Importance<\/strong> \u2013 Real world application and real world industrial setting for experimentation.<\/p>\n<p><strong>Rigour<\/strong> \u2013 Sound and rigorous methodological approach, supported by extensive experimentation<\/p>\n<h5><em><strong>Best Conference Paper Award<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Christos Koutlis \u2013 Leveraging representations from intermediate encoder-blocks for synthetic image detection<\/b><span class=\"award-citation\">C. Koutlis, S. Papadopoulos, &#8220;<i>Leveraging Representations from Intermediate Encoder-blocks for Synthetic Image Detection.<\/i>&#8220;, in Proceedings of the ECCV, 2024.<\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 The paper introduces a novel approach to synthetic image detection by leveraging intermediate representations from CLIP\u2019s image encoder. This is an interesting contribution to the field, since prior work mostly focused on final-layer embeddings. The design of the Trainable Importance Estimator adds novelty in the architecture.<\/p>\n<p><strong>Importance<\/strong> \u2013 With the exponential rise of generative AI, detecting synthetic media is a real, pressing technical challenge with important societal implications. The method achieves strong generalization across a broad range of generative models, making it a good candidate for real-world deployment in relevant systems, e.g. trust and safety ones. The algorithm outperforms state-of-the-art methods by +10.6% accuracy and trains very fast. It is important to note that the paper has already cited 23 times according to Google Scholar<\/p>\n<p><strong>Rigour<\/strong> \u2013 The experimental setup is extensive, covering 20 datasets and including comparisons with state-of-the-art methods, ablation studies, and robustness tests. The model is evaluated in multiple training configurations with clear metric reporting. The level of empirical detail is very good.<\/p>\n<h5><em><strong>Award for Best Student Journal Publication<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Thanasis Kotsiopoulos \u2013 Revolutionizing defect recognition in hard metal industry through AI explainability, human-in-the-loop approaches and cognitive mechanisms<\/b><span class=\"award-citation\">T. Kotsiopoulos, G. Papakostas, T. Vafeiadis, V. Dimitriadis, A. Nizamis, A. Bolzoni, D. Bellinati, D. Ioannidis, K. Votis, D. Tzovaras, P. Sarigiannidis, &#8220;<i>Revolutionizing defect recognition in hard metal industry through AI explainability, human-in-the-loop approaches and cognitive mechanisms.<\/i>&#8220;, Expert Systems with Applications, vol. 255, December 1, 2024.<\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 This paper presents an innovative approach to defect recognition in the hard metal industry by integrating explainable AI (XAI), human-in-the-loop (HITL) techniques, and cognitive retraining mechanisms. Unlike conventional automated defect detection systems that operate as black-box models, this study emphasizes AI transparency and human-AI collaboration. The inclusion of interpretable AI models ensures that AI-driven decisions are understandable to operators, allowing informed interventions and refinements.<\/p>\n<p><strong>Importance<\/strong> \u2013 The paper addresses key challenges in industrial defect detection, particularly trust and adaptability in AI systems. Traditional AI-based quality control methods often struggle with operator acceptance and reliability concerns, as their predictions lack intuitive explanations. By incorporating XAI, the platform provides insightful justifications for each AI decision. Furthermore, HITL mechanisms allow real-time expert feedback, enabling models to learn from human corrections and continuously improve their detection accuracy.<\/p>\n<p><strong>Rigour<\/strong> \u2013 The research presents a structured methodology, detailing model architecture, image acquisition techniques, and micro-service-based system design. The introduction of retraining mechanisms ensures AI models remain robust over time. The study uses machine and deep learning algorithms for defect classification and localization, leveraging industry-standard AI techniques. However, while simulation results are promising, further experimental validation in real data would be much desirable. Additionally, conducting a comparative analysis against existing AI-based defect detection solutions would strengthen the paper.<\/p>\n<h5><em><strong>Best Student Paper Award at a Conference<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Anestis Kastellos \u2013 FedHARM: Harmonizing Model Architectural Diversity in Federated Learning<\/b><span class=\"award-citation\">A. Kastellos, A. Psaltis, C. Z. Patrikakis, P. Daras, &#8220;<i>FedHARM: Harmonizing Model Architectural Diversity in Federated Learning.<\/i>&#8220;, in Proceedings of the European Conference on Computer Vision (ECCV 2024), Milan, Italy.<\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 Unlike standard FL approaches that require homogeneous models for weight aggregation, FedHARM focuses on harmonizing representations rather than model parameters through a hybrid training method.<\/p>\n<p><strong>Importance<\/strong> \u2013 The architecture heterogeneity problem appears to be a novel problem that this work handles in FL. As such it is an important contribution to the community although experimentation was performed in simple datasets.<\/p>\n<p><strong>Rigour<\/strong> \u2013 The methodology is clearly described, includes implementation details, and presents quantitative evaluations. The experiments are thorough across multiple datasets, architectures, and client counts. The question is how and whether the methodology adapts to other architectures, and different and more modern datasets.<\/p>\n<\/div>\n<div class=\"container\">\n<div class=\"row\">\n<div class=\"col-lg-12 col-md-12 col-12 position-title\">\n<h2 id=\"awards-2023\" class=\"award-year\"><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> <strong>Best Publications 2023<\/strong> <i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i><\/h2>\n<\/div>\n<\/div>\n<h5><em><strong>Best Journal Publication Award<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Kosmas Dimitropoulos \u2013 Multi-Manifold Attention for Vision Transformers<\/b><span class=\"award-citation\">D. Konstantinidis, I. Papastratis, K. Dimitropoulos, P. Daras, &#8220;<i>Multi-manifold attention for vision transformers.<\/i>&#8220;, In IEEE Access, doi: <a href=\"https:\/\/ieeexplore.ieee.org\/document\/10305583\" target=\"_blank\" rel=\"noopener\">10.1109\/ACCESS.2023.3329952<\/a><\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 The paper\u2019s originality is in a new self-attention mechanism that uses three different manifolds for the modelling of the input space structure. It introduces the interesting idea of distance maps being computed in each manifold and subsequently fused in an early or late fusion manner.<\/p>\n<p><strong>Importance<\/strong> \u2013 The paper has already started attracting attention from the community and its result start becoming influential. The popularity of vision transformers and the fact that the proposed method can be readily applied to any ViT, contribute to its importance and impact.<\/p>\n<p><strong>Rigour<\/strong> \u2013 The paper is well written and contains extensive experiments in two tasks (classification and segmentation) that show that the proposed approach improves the performance of ViTs where it is incorporated. Ablation studies are also included.<\/p>\n<h5><em><strong>Best Conference Paper Award<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Symeon Papadopoulos \u2013 Self-Supervised Video Similarity Learning<\/b><span class=\"award-citation\">G. Kordopatis-Zilos, G. Tolias, C. Tzelepis, I. Kompatsiaris, I. Patras, S. Papadopoulos, &#8220;<i>Self-Supervised Video Similarity Learning.<\/i>&#8220;, 2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Vancouver, BC, Canada, 2023, pp. 4756-4766, DOI: <a href=\"https:\/\/ieeexplore.ieee.org\/document\/10208782\" target=\"_blank\" rel=\"noopener\">10.1109\/CVPRW59228.2023.00504<\/a><\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 The authors propose a novel self-supervising task for video similarity using data augmentation to learn a video similarity network which then works for several downstream tasks at SoTA without using any labelling data. The approach itself is novel and interesting.<\/p>\n<p><strong>Importance<\/strong> \u2013 Self supervision has become a cornerstone of foundation models, models that perform well downstream tasks, and therefore figuring out new ways to perform SSL will dominate the future.<\/p>\n<p><strong>Rigour<\/strong> \u2013 The approach is sound, the evaluation proper, and the baselines used appropriate and rigorous.<\/p>\n<h5><em><strong>Award for Best Student Journal Publication<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> Elissavet Batziou \u2013 Artistic neural style transfer using CycleGAN and FABEMD by adaptive information selection<\/b><span class=\"award-citation\">E. Batziou, K. Ioannidis, I. Patras, S. Vrochidis, I. Kompatsiaris, &#8220;<i>Artistic neural style transfer using CycleGAN and FABEMD by adaptive information selection.<\/i>&#8220;, Pattern Recognition Letters, 165, 55-62. DOI: <a href=\"https:\/\/doi.org\/10.1016\/j.patrec.2022.11.026\" target=\"_blank\" rel=\"noopener\">https:\/\/doi.org\/10.1016\/j.patrec.2022.11.026<\/a><\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 The presented work uses a cycleGAN adapting its loss function to account for texture information. In order to obtain and transfer the right amount of spectral information the authors define an optimal number of BIMFs, which is the novelty of this work.<\/p>\n<p><strong>Importance<\/strong> \u2013 Transfer style is certainly an important and timely area of work, and this work advances the state-of-the-art in the field.<\/p>\n<p><strong>Rigour<\/strong> \u2013 The methodology followed in this work is sound. Transfer style evaluation is certainly less rigorous than typical evaluation in ML, even when compared to other generative AI applications, and as such it is always difficult to quantify the benefits of new methods.<\/p>\n<h5><em><strong>\u0392\u03c1\u03b1\u03b2\u03b5\u03af\u03bf \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7\u03c2 \u03b4\u03b7\u03bc\u03bf\u03c3\u03af\u03b5\u03c5\u03c3\u03b7\u03c2 \u03c6\u03bf\u03b9\u03c4\u03b7\u03c4\u03ae \u03c3\u03b5 \u03a3\u03c5\u03bd\u03ad\u03b4\u03c1\u03b9\u03bf<\/strong><\/em><\/h5>\n<ul>\n<li class=\"award-entry\"><b><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i> \u0394\u03b7\u03bc\u03ae\u03c4\u03c1\u03b9\u03bf\u03c2 \u0393\u03b9\u03b1\u03ba\u03bf\u03c5\u03bc\u03ae\u03c2 \u2013 Leveraging Multimodal Sensing and Topometric Mapping for Human-Like Autonomous Navigation in Complex Environments<\/b><span class=\"award-citation\">K. Tsiakas, D. Alexiou, D. Giakoumis, A. Gasteratos and D. Tzovaras, &#8220;<i>Leveraging Multimodal Sensing and Topometric Mapping for Human-Like Autonomous Navigation in Complex Environments<\/i>&#8220;, 2023 IEEE\/RSJ International Conference on Intelligent Robots and Systems (IROS), Detroit, MI, USA, 2023, pp. 7415-7421, doi: <a href=\"https:\/\/ieeexplore.ieee.org\/abstract\/document\/10341358\" target=\"_blank\" rel=\"noopener\">10.1109\/IROS55552.2023.10341358<\/a><\/span><\/li>\n<\/ul>\n<p><strong>Evaluation committee comments:<\/strong><\/p>\n<p><strong>Originality<\/strong> \u2013 The proposed method comprises original aspects such as the incorporation of common human driving attitudes in the navigation framework.<\/p>\n<p><strong>Importance<\/strong> \u2013 Autonomous driving is an area of immense activity, especially in the robotics and autonomous systems community. The proposed approach contributes to this direction by fusing information from multiple sensors, thus enabling vehicle operation in different, relatively complex environments,<\/p>\n<p><strong>Rigour<\/strong> \u2013 There is rigorous experimental evaluation of the work including an experiment involving a real vehicle.<\/p>\n<p>The process involved two independent assessments of each submitted publication by an expert and a generalist reviewer, using a four-category scale (recognised nationally \/ recognised internationally \/ internationally excellent \/ world-leading), followed by a ranking of the publications within each category.<\/p>\n<\/div>\n<div class=\"container\">\n<div class=\"row\">\n<div class=\"col-lg-12 col-md-12 col-12 position-title\">\n<h2 id=\"award-process\" class=\"award-year\"><i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i><strong>Best Paper Awards Selection Process<\/strong> <i class=\"fas fa-award fa-lg award-mark\" aria-hidden=\"true\"><\/i><\/h2>\n<\/div>\n<\/div>\n<p>With the aim of highlighting the research activities and outcomes produced at ITI, a selection process is conducted to identify the best publications for recognition through the Best Paper Awards in the following categories.<\/p>\n<p><strong>1. Awards<\/strong><\/p>\n<ol>\n<li>Best Journal Paper Award<\/li>\n<li>Best Conference Paper Award<\/li>\n<li>Best Student Journal Paper Award<\/li>\n<li>Best Student Conference Paper Award<\/li>\n<\/ol>\n<p><strong>2. Eligibility Criteria<\/strong><\/p>\n<ul class=\"award-terms\">\n<li>Publications must have been published during the year under evaluation in peer-reviewed international journals or conference proceedings.<\/li>\n<li>All co-authors must agree to the submission of the publication for consideration in the awards process.<\/li>\n<li>For all categories, the majority of the authors must be affiliated or have been affiliated with ITI (i.e., have or have had a contract with ITI).<\/li>\n<li>For categories (3) and (4), the first author must have been an active Master&#8217;s or PhD student at the time the paper was submitted.<\/li>\n<li>The subject of the publication must fall within the broader field of Information Technology and Telecommunications.<\/li>\n<li>Each ITI laboratory may submit up to three papers in total for evaluation across all categories.<\/li>\n<\/ul>\n<p><strong>3. Evaluation Committee (2023, 2024)<\/strong><\/p>\n<ul class=\"award-terms\">\n<li>Evangelos Kanoulas, University of Amsterdam<\/li>\n<li>Sokratis Katsikas, Norwegian University of Science and Technology<\/li>\n<li>Nikolaos Nikolaidis, Aristotle University of Thessaloniki<\/li>\n<li>Elpiniki Papageorgiou, University of Thessaly<\/li>\n<li>Dimitrios Pezaros, University of Glasgow (Chair)<\/li>\n<\/ul>\n<p>The process includes two independent evaluations of each submitted publication, conducted by an expert reviewer and a generalist reviewer, using a four-category scale (recognised nationally \/ recognised internationally \/ internationally excellent \/ world-leading), followed by a ranking of the papers within each category.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Best Publications 2024 Best Publications 2023 Award Selection Process Best Publications 2024 Best Journal Publication Award Apostolos Evangelidis \u2013 Efficient deep Q-learning for industrial equipment calibration in elevator manufacturingA. Evangelidis, N. Dimitriou, P. Charalampous, T. D. Mastos, D. Tzovaras, &#8220;Efficient deep Q-learning for industrial equipment calibration in elevator manufacturing.&#8220;, IEEE Transactions on Industrial Informatics, vol. [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"default","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"set","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"class_list":["post-36921","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/pages\/36921","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/comments?post=36921"}],"version-history":[{"count":4,"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/pages\/36921\/revisions"}],"predecessor-version":[{"id":36926,"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/pages\/36921\/revisions\/36926"}],"wp:attachment":[{"href":"https:\/\/www.iti.gr\/iti\/wp-json\/wp\/v2\/media?parent=36921"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}