COM4AI

Abstract: This talk explores enhancing the trustworthiness of automated driving systems through NeuroSymbolic AI. This talk propose methods where symbolic representations of qualitative spatial and temporal relationships between objects are learned from video scenes. These symbolic representations are then used as input for machine learning and pattern-mining approaches for understanding complex driving scenarios. This approach enables the generation of explanations for automated driving actions, facilitates the creation of comprehensive testing scenarios, and ultimately contributes to a more transparent and trustworthy autonomous system.

Speaker: Nassim Belmecheri, PhD, is a Research Scientist at SIMULA Research Laboratory in Oslo, Norway. He earned his Ph.D. in Computer Science from the University Oran 1 in July 2023, with his thesis focusing on "Web Data Mining: Learning from User Feedback". Prior to his current role, he served as a Postdoctoral Researcher at SIMULA. Dr. Belmecheri's expertise is at the intersection of Artificial Intelligence (AI), Data Mining, Explainable AI, and Constraint Programming. His professional experience includes developing and deploying AI solutions for companies like AXLR SATT , and working as a Data Scientist with geo-spatial data and text analysis at EvalAndGo/Qwesteo. The past 3 years he has worked on enhancing trustworthiness of AI systems in automated driving through neurosymbolic AI in the context of a European project (AI4CCAM).

Abstract: This talk presents a hardware–software co-designed intrusion detection system that leverages SmartNIC acceleration and Nested Graph Neural Networks to enable high-throughput, low-latency security analytics in modern networks.

Speaker: Rana Abu Bakar received his PhD in Emerging Digital Technologies from Scuola Superiore Sant'Anna and is working as a researcher in CNIT, Italy. Currently working on SmartNIC/DPU-accelerated networking and AI-driven security systems. His research focuses on high-speed packet processing, hardware offloading, and graph-based learning for intrusion and anomaly detection. He has published and presented work at leading IEEE venues and is actively involved in EU research projects on trustworthy and high-performance network security.

Abstract: This talk explores the use of in-network machine learning as a new paradigm for accelerating time-sensitive ML tasks. By offloading inference workloads to programmable network devices, in-network ML leverages underutilized compute resources within the data plane, achieving significant performance and energy-efficiency gains. Building on this idea, the talk presents a general methodology for prototyping, deploying, and evaluating ML inference directly inside programmable switches and NICs. Several practical applications are showcased, including ultra-low-latency stock market prediction and real-time transaction fraud detection. The talk also introduces new data structures and in-network feature-engineering algorithms tailored to resource-constrained network hardware, alongside a hybrid deployment strategy that integrates network devices with end-host servers for further acceleration. Together, these results demonstrate that in-network ML is both feasible and highly effective for time-sensitive tasks, achieving sub-microsecond inference latency and performance competitive with traditional server-based systems.

Speaker: Xinpeng Hong received his PhD in Engineering Science from the University of Oxford. His research interests lie at the intersection of networking, in-network computing, and machine learning, with a focus on time-sensitive applications of in-network machine learning. His work has been published in venues including PACMNET, ACM TODAES, ACM SIGCOMM Computer Communication Review (CCR), IEEE Communications Surveys & Tutorials (COMST), ACM CoNEXT, ECAI, and IEEE HPSR. In 2024, his research received the Prototypes for Humanity Award from the Dubai Future Foundation.