These courses aimed at providing a deeper and multi-sided understanding of a topic including hands-on experience, providing the material of the course as well as a professional development certificate. Registration to the main conference is not required to attend short courses.
The following courses will be offered at ICASSP 2026:
SC1: Tensor Decompositions for Multilayer Network Analysis, by Selin Aviyente, Borbála Hunyadi, Sofia-Eirini Kotti.
Monday, May 4, 9:00-12:30, 14:30-18:00.
Tuesday, May 5, 13:00 – 16:00.
6 hours of theoretical lectures and 3 hours of hands-on Matlab-based computer sessions.
This short course introduces tensor decompositions as a principled and scalable framework for analyzing temporal and multilayer networks. Tensors provide natural representations for high-dimensional relational data, enabling the preservation of interactions across multiple modes such as nodes, time, and layers. The course covers fundamental tensor algebra and key decomposition models, including canonical polyadic, multilinear SVD, and block term decompositions, and explains their properties, computation, and practical use. These tools are then connected to network science concepts, with a focus on community detection in multilayer networks. By establishing links between tensor generative models and classical network objectives, the course bridges modern multilinear algebra with network analysis, supported by hands-on applications to brain connectivity and social networks.
SC2: Quantum Information Processing, Sensing and Communications: Their Myths, Realities and Futures, by Lajos Hanzo
Monday, May 4, 9:00-12:30, 14:30-18:00.
6 hours of theoretical lectures
This short course provides a comprehensive introduction to quantum information processing, sensing, and communications, emphasizing both fundamental principles and emerging applications relevant to signal processing and wireless communications. Motivated by the limits of classical scaling and the transition to quantum-domain technologies, the course reviews core concepts of quantum mechanics, including superposition, entanglement, measurement, and decoherence. It introduces quantum communication models, quantum error correction, and error mitigation techniques, highlighting their connections to classical coding theory. The course further explores quantum algorithms for wireless communications, demonstrating how quantum search and optimization can address detection, estimation, localization, and routing problems. Advances in quantum key distribution, satellite-based quantum communications, and the development of quantum networks are surveyed, along with open research challenges and future direction.
SC3: Multimodal SP, Robotics and AI, by Petros Maragos, Panagiotis P. Filntisis
Monday, May 4, 9:00-12:30, 14:30-18:00.
6 hours of theoretical lectures
This short course explores the role of signal processing in modern robotics, with a focus on multimodal perception, interaction, and learning in the AI era. It addresses how heterogeneous sensory signals—including audio, visual, depth, and physiological data—can be jointly modeled, synchronized, and fused using deep neural architectures for spatio-temporal processing. The course covers multimodal human–robot interaction, emphasizing affective understanding, gesture and speech analysis, and the synthesis of expressive behaviors through 3D avatars. Applications in assistive and social robotics are discussed, highlighting human-centered perception and adaptive interaction. The course also examines advances in robot perception and learning, including self-supervised and foundation-model-based approaches, showing how signal-processing principles remain central to building intelligent, autonomous, and interactive robotic systems.