Stardust is a scalable and extensible open-source simulator for the 3D Continuum. read more

Using the YOLO11 detection framework as a starting point, various uncertainty estimation methods have been implemented and were then evaluated on several autonomous car driving datasets. read more

In my last post, I shared the results of my experiment on how model type, frame rate, and resolution affect both power consumption and prediction confidence in YOLO object detection. Those findings weren’t just for curiosity — they now form the backbone of a custom reinforcement learning (RL) environment I’ve built to explore adaptive inference on energy-constrained edge devices. read more

Last time, I introduced the setup of my experiment for measuring how model type, resolution and frame rate influence object detection performance on a RaspberryPi. Today, I’ll share results for two aspects: power consumption and prediction confidence. read more

Es ist Sommer und Halbzeit im netidee Projekt Hyper 8! Mit den zwei Releases 0.21 und 0.22 wurden wieder signifikante Meilensteine erreicht - Eine Revue der letzten Wochen. read more

Obwohl unser Projekt vom BaFG ausgenommen ist, setzen wir freiwillig auf Barrierefreiheit nach WCAG AA für ein inklusives Web. read more

At the ISVLSI 2025 conference in Kalamata, Greece, we presented our work on profiling the Qiskit compiler pipeline. Our work earned the Best Paper Award at the Quantum Computing Workshop. read more

Heute reichen wir den Endbericht ein. read more

Edge AI cuts latency to milliseconds and reduces bandwidth by up to 80 % by analyzing sensor streams locally instead of sending them to remote clouds, making it ideal for safety critical applications such as power, water and transport networks. My PhD at TU Wien combines this fast on-site inference with privacy-preserving federated learning, so sensitive data never leaves the substation while enabling swift decision making. The research impact on resilience and safety-critical applications e... read more

Seit Version 1.4 unterstützt das LAS Punktwolkenformat beliebige extra attribute. Mit Potree-Next haben wir nun auch Unterstützung für diese Eingebaut, mit besonderem Augenmerk auf Erweiterbarkeit um zahlreiche eigene Encodings. read more

Sie möchten bei netidee einreichen, haben aber noch offene Fragen? Dann sind Sie hier genau richtig! read more