Customer Reviews

A Computational Perspective on Visual Attention

5 star:		(1)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

A Computational Perspective on Visual Attention works both as an introduction to the neuroscientific study of visual attention as well as a meaningful assay of the decades-long computational vision literature.

John Tsotsos starts off by clearly outlining a broad range of computational limitations within biologically plausible neural systems and their respective restrictions on how the human visual system must operate. Through such an approach, a convincing argument is formed for the established theoretical constructs as well as the value of computational models as a whole. An analysis of the many valid approaches in the literature follows, in addition to a thorough investigation of the historically affirmed Selective Tuning model.

Overall this book is enlightening and easily read. Most importantly, perhaps, it collects and translates a fragmented literature into a common approach for the study of visual attention, and is therefore highly recommended for future students of the computational neurosciences.