“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)
Front cover of the June issue of Neurosurgery for the joint research of Taylor Frantz with UZBrussels! High-Accuracy Augmented Reality Guidance for Intracranial Drain Placement Using a Standalone Head-Worn Navigation System: First-in-Human Results
Article link : https://journals.lww.com/neurosurgery/fulltext/2025/06000/high_accuracy_augmented_reality_guidance_for.8.aspx
Journal issue link:Â https://journals.lww.com/neurosurgery/pages/currenttoc.aspx
FWO granted the project Exploiting plasma etching processes for micro/nanotexturing of metal surfaces to enable novel chemical, analytical, optical, and medical applications.
The plasma metal etcher will be installed in the core facility MICROLAB. The project execution will be coordinated by Prof. Wim de Malsche (CHIS) with two ETRO-promoters Prof. Johan Stiens and Prof. Peter Schelkens. Step by step the microfabrication facilities are growing, allowing to dive deeper again in microfabrication related projects.
On October 6th 2025 at 16:00, Ran Zhao will defend their PhD entitled “DEEP LEARNING-BASED HUMAN POSTURE NORMALIZATION AND AUTOMATIC ANTHROPOMETRIC MEASUREMENT”.
Everybody is invited to attend the presentation in room D.2.01 or online via this link.
Accurate and user-friendly anthropometric measurement remains a major challenge in computer vision, as existing approaches typically require controlled scanning conditions, standard postures, or unclothed bodies. These constraints limit their usability in practical scenarios.
This thesis proposes a sequence of deep learning-based solutions to overcome these limitations. We first introduce OrienNormNet, an iterative network for robust orientation normalization, ensuring that scans are consistently aligned without manual preprocessing. Building on this, PoseNormNet is presented as the first posture normalization framework that transforms arbitrarily posed scans into a canonical T-pose while preserving identity details, removing the need for skeleton rigging. Next, W2H-Net demonstrates the feasibility of directly estimating the waist-to-hip ratio from partial dressed scans, showing that reliable indicators can be derived even from incomplete data. Finally, MeasureXpert provides a breakthrough toward real-world usability: it enables automatic extraction of anthropometric measurements from only two unregistered, partial, and clothed scans acquired in arbitrary poses.
To support these developments, the BWM dataset was synthesized for training, validation, and evaluation. Comprehensive experiments on both synthetic and real-world data confirm the effectiveness and robustness of the proposed methods. Collectively, the contributions progressively address key challenges related to cost, posture, and clothing, moving the field closer to practical, flexible, and accessible body measurement solutions.
The algorithms presented in this thesis have been disseminated through prestigious journals and conferences, demonstrating a modest yet meaningful impact on both academic research and industrial applications.
The objective of the Surv-AI-llance consortium was to build a reliable, privacy-friendly video analytics pipeline that accurately interprets surveillance scenes and rapidly alerts officials. During the closing event, the researchers present their findings, while the industry partners present their valorization plans and how to enable A.I. in the police-force workspaces.
Register here for the closing event: Â https://drive.google.com/file/d/1bwli-pAjs0tYCzTkO6Up-pd7zWfe9SvI/view?usp=drive_link
More info on the project: www.surv-ai-llance.com
23 February 2024
Corda Campus, building 1, floor 1
Kempische Steenweg 293/16, 3500 Hasselt
Collapse of the Soviet Union – a disaster for Russian science and research, an opportunity for structural collaboration on hitherto unexplored theories (mm waves 30- 300 GHz).
On June 21 2023 at 16.00, Nicolas Ospitia Patino will defend his PhD entitled “UNRAVELING TEXTILE-REINFORCED CEMENTITIOUS COMPOSITES BY MEANS OF MULTIMODAL SENSING TECHNIQUES”.
Everybody is invited to attend the presentation at the Room D.0.08.
.
Textile Reinforced Cementitious (TRC) sandwich composites are innovative construction materials composed of two slender TRC facings, and a thick thermal and acoustic insulating core. Their non-corrosive nature allows for slender structures, resulting in a reduction of the cement used, and therefore a decrease of the negative impact on the environment. The sandwich technology brings superior bending resistance while enforcing the lightweight nature of the composite. Despite the numerous advantages of TRC sandwich composites, they present a complex and possibly unpredictable fracture behavior, and manufacturing issues such as a weak interlaminar bond and therefore, need status verification in the different stages of their service life: at manufacturing stage (curing), final product quality (manufacturing defects), deterioration during use (damage accumulation). Up to the moment, there is no reliable non-invasive inspection protocol that assesses the curing of the cementitious facings, provide quality control, and damage monitoring.
Along this study, a combination of Non-Destructive Testing (NDT) techniques is employed to provide a protocol that allows to monitor the composite from the hardening of the cementitious facings, quality control, and finally, damage characterization. Electromagnetic millimeter wave (MMW) spectrometry is employed for the first time in this kind of material to monitor the hydration of cementitious media, quality control, and damage characterization. Additionally, passive, and active elastic wave-based NDT techniques, like Acoustic Emission (AE) and Ultrasound, respectively, are also used in combination with Digital Image Correlation(DIC) to characterize the material along its lifetime, and benchmark MMW spectrometry. This thesis summarizes the results of an extensive experimental campaign, highlighting the innovative contributions. Previously unknown relations between electromagnetic properties measured by MMW and mechanical properties by ultrasound are revealed owing to the common origin of hydration reaction that dictates the permittivity and stiffness development. AE during proofloading reveals the effect of manufacturing defects due to the local stress field variations they impose under mechanical test. In addition, cracking and debonding leave a strong fingerprint on the electromagnetic transmission, enabling a multi-spectral methodology for the structural health monitoring (SHM) of such innovative components during their lifetime.