Kuva: Eetu Soronen
Small edge devices, such as drones, often offload demanding computer vision tasks to a nearby server because they lack computing power. This offloading relies on wireless transmission, which adds delay, consumes power and introduces unpredictability. Compression reduces transmission time and energy, but also requires computing power and may degrade image quality. In his research, Ing. Jakub Žádník tackles these challenges through low-complexity image compression, joint source-channel coding optimized for high accuracy in noisy channels, and adaptive selection of the best compression strategy based on network quality. Together, these results help edge devices run smarter vision applications even in changing network conditions.
The doctoral dissertation of Ing. Jakub Žádník in the field of computer engineering titled Edge Offloading of Low-Latency Computer Vision Tasks in Challenging Network Conditions will be publicly examined at the Faculty of Information Technology and Communication Sciences at Tampere University on 31 March 2026.
The Opponent will be Dr. Rajesh Sankaran from the Argonne National Laboratory in the United States of America. The Custos will be Prof. Pekka Jääskeläinen from the Faculty of Information Technology and Communication Sciences, Tampere University.