“Signal Processing in the AI era” was the tagline of this year’s IEEE International Conference on Acoustics, Speech and Signal Processing, taking place in Rhodes, Greece.
In this context, Brent de Weerdt, Xiangyu Yang, Boris Joukovsky, Alex Stergiou and Nikos Deligiannis presented ETRO’s research during poster sessions and oral presentations, with novel ways to process and understand graph, video, and audio data. Nikos Deligiannis chaired a session on Graph Deep Learning, attended the IEEE T-IP Editorial Board Meeting, and had the opportunity to meet with collaborators from the VUB-Duke-Ugent-UCL joint lab.
Featured articles:
- De Weerdt, B. et al., “Designing Transformer Networks for Sparse Recovery of Sequential Data Using Deep Unfolding” (https://ieeexplore.ieee.org/abstract/document/10094712)
- Yang X. et al., “Relevance Propagation through Deep Conditional Random Fields” (https://ieeexplore.ieee.org/abstract/document/10095075)
- Stergiou A. et al., “Play it back: Iterative attention for audio recognition” (https://ieeexplore.ieee.org/abstract/document/10096532, developed at the University of Bristol)
VUB, ULB and the Brussels Capital Region government announce the launch of a new Artificial Intelligence for the Common Good Institute during the Belgian AI Week of AI4Belgium.
Today, Annemie Schaus and Caroline Pauwels, respectively rectors of ULB and VUB, were proud to announce the launch of FARI the Artificial Intelligence institute for the Common Good. It was also endorsed by the Brussels Capital Region ministers Barbara Trachte and Bernard Clerfayt.
FARI is a unique structure that aims at gathering over 300 researchers in AI (amongst others the ETRO dept) and associated disciplines, around projects that could benefit the general interest. The institute will promote research on trustworthy, transparent and explainable artificial intelligence. It will also aim at helping the Brussels Region and its inhabitants address some of the challenges they face in various domains. FARI researchers will provide ideas and contribute to projects on transportation, sustainable development, healthcare services, civic consultations on AI and algorithms. Its projects will actively involve citizens and reinforce education on AI and its impacts in the region.
The institute aims at creating a bridge between AI experts, citizens, companies and local organizations. It will have three hubs: an Research & Innovation Hub, a Think Tank on AI, Data and Society, and an AI Test and Experience Hub.
https://today.vub.be/en/article/fari-a-new-artificial-intelligence-institute-in-brussels
ETRO VUB, developed a radar that makes it possible to look in and through walls, ceilings and floors. The instrument is intended to help renovators and restorers in buildings where a detailed technical plan is lacking.
In many buildings it is still too often guessing what is in the walls. Moreover, you don’t just throw open a floor or ceiling.
The radar can now calculate the exact thickness of walls, the size of pipes that run through them, can map cavities, so that as a renovator you no longer have to work blindly.
A classic radar emits electromagnetic waves that contain many frequencies, but that leads to a complex signal in which no materials or their thickness can be deduced. The ETRO radar uses the material-dependent propagation speed and the reflection of electromagnetic waves.
Using complex mathematical models, the cross-section of the wall with everything in it can be made from the reflected image.
For the construction sector, this offers possibilities such as detecting insulation materials or leaks in walls and floors or detecting invisible concrete rot, for the professionalization of restoration projects.
“It can be used in stone, wood, glass, concrete and can even separate layers of paint with a thickness of 100 micrometers. It does not work for metals, but of course you see them appearing in the image as obstacles.”
https://www.bruzz.be/wetenschap/nieuwe-radar-van-vub-team-kijkt-en-door-muren-2022-05-27
https://time.news/new-radar-from-vub-team-looks-in-and-through-walls/
https://www.notiulti.com/el-nuevo-radar-del-equipo-vub-mira-dentro-y-a-traves-de-las-paredes/
https://www.hln.be/brussel/onderzoeksteam-vub-kan-met-radar-door-muren-kijken~a589cdf6/
https://www.vub.be/events/2022/nieuwe-vub-radar-om-in-de-muren-te-kijken
https://press.vub.ac.be/nieuwe-vub-radar-om-in-de-muren-te-kijken
https://www.archytele.com/new-radar-from-vub-team-looks-in-and-through-walls/
https://drimble.nl/dossiers/wetenschap/84702510/vub-onderzoeksteam-ontwikkelt-murenradar.html
https://nl.metrotime.be/bizar/zotjes-nieuwe-radar-van-vub-team-kijkt-en-door-muren
https://engineeringnet.be/nl/nieuws/item/19961
Abel DĂaz Berenguer (Cuba) joined ETRO in 2017 and obtained his PhD in 2021. His father was a Civil engineer, and during his childhood, he spent a lot of time with him in construction works. This triggered his curiosity to build and create things. Since Abel was a kid, he wanted to study ¨something¨ with computers and never had any doubt about studying engineering in informatics sciences.
The PhD program made him feel thoroughly responsible for your research project. Advisors progressively introduce you into the research environment and educate you on digging deeper into fundamental theories by promoting critical thinking. Abel noticed a lack of motivation to participate in dissemination activities that promote science communication. Students also did not feel the need to communicate and to encourage more collaboration between other fellows working in the same or different fields.
Abel enjoyed the Writing Bootcamp of the doctoral Training program very much. The three days course about writing scientific articles offered mainly many good tips about approaching the scientific writing process, which was a great help during the PhD.
Abel has great memories of chats, coffees, and plenty of amusing moments with colleagues. Those moments allowed him to overcome challenging days. The collaboration with close colleagues has been outstanding. Research is a process of knowledge sharing and co-creation with excellent colleagues that became family that stood by my side on long working days and nights.
Abel grew during his PhD into a person with more critical thinking towards solving any problem in life. Be enthusiastic and passionate about your research. Enjoy the learning process with perseverance, dedication, engagement, and curiosity for innovating. Push boundaries and never give up share knowledge and be a team player.
Abel wants to land in an academic environment, help others learn and learn from them. In a position to benefit society, share knowledge and contribute to building a better world for our children.
– The language of the programme is English, and the requirements are described here: http://www.vub.ac.be/en/study/applied-sciences-and-engineering-applied-computer-science#admission-criteria;
– If cannot provide a proof of sufficient knowledge of English, then, after the positive evaluation of your academic background and all other criteria to follow the program are met, through an interview the professors will also assess your level of English and a final decision will be taken.
For scholarship opportunities you can check our website: http://www.vub.ac.be/en/study/applied-sciences-and-engineering-applied-computer-science#practical-info
All students who apply for this program are eligible to be awarded with a VUB Scholarship. The scholarships will be awarded on a competitive basis among all accepted applicants.
On October 25th 2024 at 16:00, Yuqing Yang will defend their PhD entitled “CRAFTING EFFECTIVE VISUAL EXPLANATIONS BY ATTRIBUTING THE IMPACT OF DATASETS, ARCHITECTURES AND DATA COMPRESSION TECHNIQUES”.
Everybody is invited to attend the presentation in room D.2.01 or online via this link.
Explainable Artificial Intelligence (XAI) plays an important role in modern AI research, motivated by the desire for transparency and interpretability within AI-driven decision-making. As AI systems become more advanced and complicated, it becomes increasingly important to ensure they are reliable, responsible, and ethical. These imperatives are particularly acute in domains where stakes are high, such as medical diagnostics, autonomous driving, and security frameworks.
In computer vision, XAI aims to provide understandable, straightforward explanations for AI model predictions, allowing users to grasp the decision-making processes of these complex systems. Visualizations such as saliency maps are frequently employed to identify input data regions significantly impacting model predictions, thus enhancing user understanding of AI visual data analysis. However, there are still concerns about the effectiveness of visual explanations, especially regarding their robustness, trustworthiness, and human-friendliness.
Our research aims to advance this field by evaluating how various factors—such as the diversity of datasets, the architecture of models, and techniques for data compression—influence the effectiveness of visual explanations in AI applications. Through thorough analysis and careful refinement, we strive to enhance these explanations, ensuring they are both highly informative and accessible to users in diverse XAI applications.
During our evaluation process, we conduct a detailed investigation using both automatic metrics and subjective evaluation methods to assess the effectiveness of visual explanations thoroughly. Automatic metrics, such as task performance and localization accuracy, provide quantifiable measures of the effectiveness of these explanations in real-world scenarios. For subjective evaluation, we have developed a framework named SNIPPET, which enables a detailed and user-oriented assessment of visual explanations. Additionally, our research explores how these objective metrics correlate with subjective human judgments, aiming to integrate quantitative data with the more nuanced, qualitative feedback from users. Ultimately, our goal is to provide comprehensive insights into the practical aspects of XAI methodologies, particularly focusing on their implementation in the field of computer vision.