Sidi Ahmed Mahmoudi & Mohamed Benkedadra at Hannover Messe 2023 in Germany 🇩🇪 with WBI/AWEX
DeepILIA showcased its innovative AI-based worksite security solution, InfraSecure, at Hannover Messe, Europe's largest industrial technology fair. Presented by Pr. Sidi-Ahmed Mahmoudi and Ir. Mohamed Benkedadra. the solution, highlighted the lab’s commitment to enhancing worker safety through AI and attracted significant international interest, demonstrating the versatility of AI for applications in various sectors, including production lines and public spaces.
Sidi Ahmed Mahmoudi & Mohamed Benkedadra at Hannover Messe 2023 in Germany 🇩🇪 with WBI/AWEX
DeepILIA showcased its innovative AI-based worksite security solution, InfraSecure, at Hannover Messe, Europe's largest industrial technology fair. Presented by Pr. Sidi-Ahmed Mahmoudi and Ir. Mohamed Benkedadra. the solution, highlighted the lab’s commitment to enhancing worker safety through AI and attracted significant international interest, demonstrating the versatility of AI for applications in various sectors, including production lines and public spaces.
Mohamed Benkedadra, UMONS UCL, ULB, and Multitel in Nantes 🇫🇷 for Trail SW 2023 to Create CIA, a Framework for Data Augmentation Using Stable Diffusion 🚀
Mohamed Benkedadra, collaborated with researchers from UCL, ULB, and Multitel at the TRAIL Summer Workshop 2024 in Nantes. The team developed the CIA framework using Stable Diffusion and ControlNet, enhancing data augmentation for applications like object detection and medical imaging. Their work was published at the MIPR 2024 conference. The workshop provided a dynamic environment for networking and future collaborations.
Otmane Amel and Sedrick Stassin in ESANN 2023 to present Multi Modal AI for Customs
Otmane Amel and Sédrick Stassin, researchers in multimodal learning and explainability, presented their research at ESANN 2023, held from October 4-6. They published two papers: one on using multimodal deep learning for customs code prediction, achieving impressive accuracy with a novel fusion method, and another comparing semantic similarity and supervised models for HS code prediction. The event not only provided a platform for collaboration but also ended with a visit to the beautiful city of Bruges !
Sédrick Stassin Leads Collaborative Research on Bias Detection at Trail Summer Workshop 2022 at the Belgian Embassy in Berlin, Germany 🇧🇪🇩🇪
Sédrick Stassin and his team presented their project on bias detection and mitigation at the TRAIL Summer Workshop 2022 in Berlin. The research focuses on improving AI model fairness by detecting hidden biases in unstructured data like images and text, providing essential tools for ethical AI development.
Rahmani Taha Abdelaziz Presents Doctoral Thesis on Intelligent and Distributed Resource Management in Heterogeneous Cloud Computing Systems 🇩🇿
Prof. Sidi Ahmed Mahmoudi as Co-Director for Rahmani Taha Abdelaziz who defended his doctoral thesis on intelligent and distributed resource management in heterogeneous (CPU/GPU) cloud computing systems, focusing on optimizing resource use for better performance and scalability.
Zainab Ouardirhi Presents a New Approach to Occluded Object Detection at CloudTech'23 in Marrakesh 🇲🇦
Zainab Ouardirhi participated in CloudTech'23 as an author and organizing committee member, presenting her innovative research on enhancing object detection in smart video surveillance systems. Her paper introduces a novel approach using Feature Pyramid Networks and point cloud density analysis to address occlusion challenges, marking a significant advancement in AI and computer vision.
Mohamed Benkedadra Showcases AI Safety Solution at "Ma Thèse en 180 Secondes" 2024 🚧🤖
Mohamed Benkedadra presented his AI-driven safety solution at "Ma Thèse en 180 Secondes" 2024, using 3D cameras to detect dangers on construction sites and protect workers in real-time. His research emphasizes enhancing safety while respecting privacy.
Sidi Ahmed Mahmoudi & Mohamed Benkedadra with WBI/AWEX at Hannover Messe 2024 🇩🇪 Presenting the INFRASECURE AI Box
UMONS showcased the InfraSecure Surveillance Box at Hannover Messe 2024, highlighting advanced AI-driven safety solutions for construction sites and public environments. The innovative technology, presented by researchers Sidi-Ahmed Mahmoudi and Mohamed Benkedadra, drew significant interest for its potential to enhance safety and operational efficiency across various industries.
Sédrick Stassin and Sidi Ahmed Mahmoudi at ICCV 2023 in Paris 🇫🇷
Sédrick Stassin, along with his PhD advisor Prof. Sidi Ahmed Mahmoudi, attended ICCV 2023, for a paper they co-authored. This paper, part of the TRAIL initiative, aims to make AI models more transparent and trustworthy, addressing the challenges of AI's “black box” nature in critical applications.
UMONS DeepILIA Researchers Shine at TRAIL Summer Workshop 2024 in Lisbon, Portugal 🇵🇹🌞
UMONS researchers Mohamed Benkedadra, Sédrick Stassin, and Tojo Valisoa Johanesa worked on innovative AI projects at the TRAIL Summer Workshop 2024 in Lisbon, focusing on synthetic data augmentation, explainable AI, and energy consumption prediction. Their work highlights AI’s potential to address real-world challenges and contribute to smarter, more sustainable solutions.
Mohamed Benkedadra Presents a TRAIL Research paper on Diffusion Models at IEEE MIPR 2024 in San Jose, California 🇺🇸
Mohamed Benkedadra presented the CIA framework for Data Augmentation at IEEE MIPR 2024 in San Jose, showcasing how Stable Diffusion and ControlNet can enhance synthetic data generation for AI applications. The conference offered valuable networking opportunities and concluded with a scenic tour around the San Francisco area.
Otmane Amel's INFORTECH Presentation: Multi-Modal AI from Customs Fraud Detection to Railway Construction Sites 🚧
Otmane Amel presented his innovative work on Multi-Modal AI at INFORTECH, showcasing how combining data from multiple sources improves AI accuracy in real-world applications like customs fraud detection and railway construction site safety. His MutlConcat Fusion method, which preserves modality-specific features while extracting cross-modal insights, proved key in enhancing model performance.
UMONS at InnoTrans 2024 in Berlin 🇩🇪 With AWEX/WBI and INFRABEL
UMONS showcased its AI-driven safety solutions at InnoTrans 2024, highlighting projects like InfraSecure and fostering partnerships, including discussions with Infrabel's CEO. This participation emphasized Belgium’s leadership in technological innovation for railway and industrial safety.
We publish code for researchers and for industry !
We have published multiple frameworks and libraries for the research community, some were created during workshops or as part of publications. Others, are the results of the work done in multiple student project at UMONS. We are also proud to have worked on multiple industrial projects with amazing partners, who allowed us to share the knowledge.