Andrew Kiruluta: Optical Intelligence, Gebunden

Klappentext

Optical Intelligence develops a rigorous and wide-ranging theory of how optics and artificial intelligence can be understood within a single mathematical and physical framework. The book begins from first principles, treating light as a structured medium of information transformation governed by Maxwell's equations, wave propagation, Fourier analysis, statistical optics, and inverse problems. From there, it shows how classical optical processing, computational imaging, and functional analysis naturally connect to core ideas in modern AI, including representation learning, operator composition, kernel methods, optimization, and inference. Its central thesis is that optics is not merely a sensor front-end for digital intelligence, but a physically grounded computational substrate whose geometry, propagation laws, and material structure can actively shape learning and decision-making.

As the book progresses, it moves from foundations into advanced architectures and emerging research directions, including diffractive neural networks, integrated photonic systems, nonlinear optical learning, optical reservoirs, holographic memory, and hybrid optical-electronic intelligence. Throughout, the text emphasizes both theoretical depth and systems-level insight, showing how physical propagation, measurement, and learning can be co-designed to create new forms of intelligent sensing and computation. The result is a unified account of optical intelligence as a serious scientific discipline at the intersection of electromagnetism, signal processing, machine learning, and information theory. Rather than presenting optics and AI as separate domains that occasionally interact, the book argues that their deepest future lies in their integration into trainable, physically embodied systems for imaging, inference, communication, and adaptive computation.