TwitterLinkedInGitHubMediumYouTubegoogle scholarLinkLinkLinkLink
Link

I am a Research Lecturer at the KIOS Research & Innovation Center of Excellence at the University of Cyprus. I earned my Ph.D. in Computer Engineering from the Department of Electrical and Computer Engineering at the University of Cyprus in 2014. I graduated top of my class with a B.Sc. in Computer Engineering in 2008 and completed an M.Sc. in Computer Engineering in 2010 with a full scholarship, all from the University of Cyprus.

My research lies at the intersection of Computing, Machine Learning, and Computer Vision. I have made significant contributions in the areas of computation/data-efficient machine/deep learning design/search (Cascade SVM), and perception learning for vision systems (EmergencyNet, DroNet), particularly in recognition, detection, segmentation, and control. My work has resulted in over 65 scientific publications, including two best paper awards. My long-term research goal is to address the questions: a) How can we build learning machines that make efficient use of available resources (in terms of compute and available data)? and b) How learning machines can dynamically adapt based on context to improve efficiency and robustness? 

My expertise extends to actively participating in the conception, design, and implementation of numerous National and European funded research and innovation projects. My research activities encompass applications in emergency management, traffic monitoring, autonomous vehicles, and surveillance. I also serve as a reviewer for several computer vision and machine learning journals (e.g., IEEE Trans. PAMI, IEEE Trans. Artificial Intelligence; IEEE Trans. Neural Networks and Learning Systems, IEEE Trans. Image Processing), and conferences (e.g., CVPR, ECCV, ICCV, WACV,  NeurIPS, ICML, ICLR). 

NEWS

JUL 2024📜Paper Accepted at IEEE TAI - Spatiotemporal Object Detection for Improved Aerial Vehicle Detection in Traffic Monitoring

JUL 2024🗣Paper Presented at 5th International Conference on Deep Learning Theory and Applications (DeLTA) - "Closing the Sim-to-Real Gap: Enhancing Autonomous Precision Landing of UAVs with Detection-Informed Deep Reinforcement Learning" - presentation

JUN 2024📜Paper Accepted at SN Computer Science - DiRecNetV2: A Transformer-Enhanced Network for Aerial Disaster Recognition

APR 2024📜Paper Accepted at DeLTA2024 - Closing the Sim-to-Real Gap: Enhancing Autonomous Precision Landing of UAVs with Detection-Informed Deep Reinforcement Learning

MAR 2024 🥇 New Funded Proposal under Horizon Europe - GuardAI (~€5M)

MAR 2024📜Paper Accepted at IEEE TNNLS - Towards Efficient Convolutional Neural Networks with Structured Ternary Patterns

FEB 2024 🗣  New paper presented at AAAI - Convolutional Channel-wise Competitive Learning for the Forward-Forward Algorithm - arxiv

Edge Vision and AI 

"My research centers around machine learning and computer vision with a focus on developing efficient and robust representation learning algorithms for intelligent systems that are able to operate in dynamic and uncertain environments, perceive the world through multiple modalities in an adaptive fashion, and autonomously act in it in a way that both trustworthy and safe for humans."

EmergencyNet: Efficient Aerial Image Classification for Drone-Based Emergency Monitoring

YOLOpeds: Efficient real-time single-shot pedestrian detection for smart camera applications

C^3 Net: End-to-end deep learning for active smart camera control 

AirCamRTM: Enhancing Vehicle Detection for Efficient Aerial Camera-based Road Traffic Monitoring

HIGHLIGHTED ARTICLES

Machine Learning for Emergency Management: A Survey and Future Outlook

The article surveys the use of machine learning in emergency management, highlighting its role in handling data from natural and human-made disasters. It categorizes relevant works and outlines challenges and future directions, emphasizing the need for improved algorithm generalization and transparency for emergency personnel.

PROCEEDINGS IEEE - 2023

How High can you Detect? Improved accuracy and efficiency at varying altitudes for Aerial Vehicle Detection

This paper explores the challenges of object detection in aerial images, particularly regarding small objects and varying altitudes, proposing a solution using a single CNN model trained on mixed-altitude datasets to achieve high accuracy and real-time processing.

ICUAS 2023

UPCOMING EVENTS

CONTACT ME

KIOS Research and Innovation Center of Excellence

1 Panepistimiou Avenue, 2109 Aglantzia,

Tel: (+357) 22 893450 / 22 893451

Email: ckyrkou@gmail.com

Email: kyrkou.christos@ucy.ac.cy