Hamid Sheikh, Vice President, Samsung Research

This presentation addresses the continuous integration and evolution of artificial intelligence in signal processing within consumer electronics. Presented by Dr. Hamid Sheikh, Vice President at Samsung Research, for the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2026), the primary objective is to examine how modern AI techniques—particularly generative AI—are redefining consumer experiences across cameras, audio systems, computer vision, and health devices, while navigating the severe physical and computational boundaries imposed by mobile hardware.

To achieve this, the address details the technical friction between rapidly escalating consumer expectations and the uncompromising limitations of small-form-factor devices. Key points center on the algorithmic adaptations required to process complex signal modalities under strict hardware constraints, including low light, degraded signal strength, limited thermal dissipation, finite processing throughput, and small battery capacities. Main evidence emphasizes that scaling high-complexity generative and deep-learning models down to edge devices demands targeted algorithmic optimizations, hardware-aware execution, and lightweight inference architectures tailored for real-time sensing and media processing.

The ultimate takeaway is that the future of consumer signal processing relies on bridging the gap between state-of-the-art AI performance and on-device execution efficiency. By treating tight hardware constraints as design drivers rather than mere roadblocks, modern signal processing can successfully deliver immersive, responsive, and intelligent user experiences right at the mobile edge.