Stefanos Laskaridis

About

Stefanos Laskaridis is a researcher specializing in efficient deep learning, mobile vision systems, and neural network optimization, with a particular focus on making sophisticated computer vision accessible on resource-constrained devices. His work addresses one of the central challenges in modern AI deployment: enabling complex visual understanding tasks to run effectively within the tight latency and computational budgets of embedded and mobile hardware. Laskaridis is best known for his pioneering contributions to multi-exit neural network architectures applied to semantic segmentation — a fundamental capability underlying self-driving vehicles, augmented reality, robot navigation, and teleconferencing. His landmark paper "Multi-Exit Semantic Segmentation Networks," which has accumulated 24 citations, introduces adaptive inference mechanisms that allow networks to exit computation early depending on available resources, offering a compelling trade-off between accuracy and efficiency without redesigning models from scratch. His broader research agenda, reflected in work on adaptable mobile vision systems, demonstrates a consistent commitment to bridging the gap between cutting-edge deep learning and real-world deployment constraints. For students and researchers working at the intersection of computer vision and systems optimization, Laskaridis's contributions offer both practical frameworks and conceptual clarity for building leaner, smarter vision pipelines.

Research Focus

Key Achievements

Top Papers

1. 1
   Multi-Exit Semantic Segmentation Networks

   24 citations · 2022
2. 2
   Adaptable mobile vision systems through multi-exit neural networks

   3 citations · 2022
3. 3
   Multi-Exit Semantic Segmentation Networks

   2 citations · 2021

Key Collaborators

Contact & Links