A one-hour webinar
From Parts to Whole: Capsule Networks for Interpretable and Transparent AI
15 jan. 2026 11:00 - 12:00
Get familiar with how Capsule Networks work, why they matter, and how they bring us closer to truly interpretable AI
Praktische info:
15 jan. 2026 11:00 - 12:00
1 uur
MS
Engels
Doelgroep: Professionals and Academic Individuals working on Artificial Intelligence and Machine Learning related problems.
Inschrijven?
- Voorwaarden: Familiarity with Deep Neural Networks
- Prijs: Free of Charge
georganiseerd door:
Leertraject
Are you curious about what lies beyond traditional Deep Convolutional Neural Networks (CNNs)? Deep learning has changed how machines perceive the world, but traditional CNNs often miss the bigger picture. They recognize what is present, yet struggle to understand how parts relate to a whole. Capsule Networks (CapsNets) offer a fresh perspective: a neural architecture that captures spatial hierarchies and relationships, which provides more transparent and human-understandable representations
Join us to explore how Capsule Networks work, why they matter, and how they bring us closer to truly interpretable AI. Discover their advantages over CNNs, their role in Explainable AI, and the growing research shaping the future of transparent deep learning.
Lesgever/spreker
Saja Tawalbeh
Saja Tawalbeh received her M.S. degree in Computer Science (Natural Language Processing) from Jordan University of Science and Technology, Irbid, Jordan, in 2020. From 2018 to 2021, she was a research assistant in NLP at the same university. She is currently a senior research fellow and pursuing her Ph.D. in Explainable Artificial Intelligence (XAI) at the Faculty of Science, University of Antwerp, sqIRL/IDLab, imec, Antwerp, Belgium. Her research interests include NLP, Model Interpretability and Explainability, with a focus on Capsule Networks and Convolutional Neural Networks (CNNs) for both image and text-based analysis.
Gerelateerde opleidingen
