Research Focus

I work at the intersection of continual learning, neuromorphic computing, and embedded AI. My goal is to enable autonomous agents (robots, drones, wearables) to learn new tasks on the fly, from streaming sensor data, under tight latency and energy budgets. To this end, I co-design online learning algorithms with the neuromorphic hardware they run on (notably Intel’s Loihi 2 chip), focusing on prototype-based and spiking neural networks with local plasticity rules. Recent work targets few-shot class-incremental learning, event-camera perception, and action recognition for robotics.

Selected Publications

  • Elvin Hajizada, Danielle Rager, Timothy Shea, Leobardo Campos-Macias, Andreas Wild, Eyke Hüllermeier, Yulia Sandamirskaya, Mike Davies. Online Continual Learning on Intel Loihi 2 via a Co-designed Spiking Neural Network. Submitted 2025, under revision at Nature Communications (2026). arXiv: 2511.01553. DOI: 10.48550/arXiv.2511.01553.
  • Elvin Hajizada†, Michael Neumeier†, Edward Paxon Frady, Yulia Sandamirskaya, Axel von Arnim, Bing Li, Eyke Hüllermeier. CLANE: Continual Learning of Actions on Neuromorphic Hardware from Event Cameras. Proc. ICANN 2026. †Equal contribution. arXiv: 2605.28387. DOI: 10.48550/arXiv.2605.28387.
  • Elvin Hajizada, Balachandran Swaminathan, Yulia Sandamirskaya. Continual Learning for Autonomous Robots: A Prototype-based Approach. Proc. IROS 2024. DOI: 10.1109/IROS58592.2024.10802683.
  • Elvin Hajizada, Patrick Berggold, Massimiliano Iacono, Arren Glover, Yulia Sandamirskaya. Interactive Continual Learning for Robots: A Neuromorphic Approach. Proc. ICONS 2022. Best Paper Award. DOI: 10.1145/3546790.3546791
.