Darko Jelaca

Papers

Total Citations

H-Index

About

Darko Jelaca is a leading researcher in bio-inspired vision systems and hardware acceleration for neural networks, with a focus on dataflow architectures for real-time visual processing. His most notable contribution is the development of the **NeuFlow system-on-a-chip**, a groundbreaking dataflow vision processing platform implemented in IBM’s 45 nm SOI process. This work, which has garnered **over 86 citations**, introduced a novel approach to accelerating complex vision algorithms—particularly those relying on large-scale convolutions and matrix operations—by mimicking the parallel, event-driven processing of biological neural networks. Jelaca’s research bridges the gap between algorithmic efficiency and hardware implementation, enabling high-speed, low-power vision systems for applications ranging from autonomous navigation to medical imaging. His achievements include pioneering the integration of neural network accelerators into compact, energy-efficient SoCs, a key step toward deploying advanced computer vision in embedded and mobile devices. By demonstrating that dataflow architectures can rival traditional GPU-based systems in performance while consuming a fraction of the power, Jelaca has established himself as a pivotal figure in the evolution of neuromorphic computing and real-time visual intelligence.