Kreiman, G. (2021). Biological and Computer Vision

This is a brilliantly inspiring, and highly accessible, work by Gabriel Kreiman (Centre for Brains, Minds, and Machines, Harvard Medical School). Kreiman is an accomplished scientist and past recipient of the National Institute of Health (NIH) New Investigator Award, National Science Foundation (NSF) Career Award, Pisart Award for Vision Science, and the McKnight Foundation Research Award. He is also a talented writer and communicator. The reader is in good hands. As Kreiman describes in the Preface, Biological and Computer Vision grew out of his research-led teaching on biological and computer vision to undergraduate and postgraduate students arriving from a disparate variety of backgrounds spanning psychology, neuroscience, computer vision, molecular biology, mathematics, engineering, and philosophy. One can imagine the scale of the challenge. Kreiman notes that his goal in the book is to whet the student’s appetite for research in vision. He achieves this admirably and more besides. But while it may be aimed primarily at the student reader, it is also highly recommended for seasoned veterans who themselves may just be starting to consider how their own work may be enhanced, and inspired, by cross-fertilisation of ideas between the biological and machine vision domains. Essential reading for students and researchers. This work brilliantly conveys the exciting and limitless potential (and challenges) of an interdisciplinary approach to understanding vision. Rather than attempting to cover every topic in the field, Kreiman introduces the main issues, findings, and debates across a selection of key areas—avoiding also to include overly long lists of citations and references (which are restricted to five key readings per chapter— although more extensive reference lists can be consulted in the online supporting materials which also usefully include other links and video material). This was a good choice as it facilitates the flow of the text and maintains engagement, and fascination, from beginning to end. Each chapter concludes with a useful bullet point summary of the main take home messages and key points. The first half of the book covers the basics of low-level visual neurophysiology, natural image statistics, light, the retina and sensory transduction, perceptual organisation, and visual illusions before moving on to midand higher-level vision including lessons learned from studies of brain lesions and brain stimulation, neural circuits and tuning preferences, Perception