Toshifumi Kato

Hewlett-Packard (Japan), École Polytechnique Fédérale de Lausanne

Papers

Total Citations

119

H-Index

About

Toshifumi Kato is a pioneering researcher in evolutionary robotics and active vision systems. His work centers on the coevolution of visual perception and selective attention, exploring how artificial agents can autonomously develop efficient scanning strategies to interpret complex visual scenes. Kato's most influential contribution, "Coevolution of Active Vision and Feature Selection" (2004, 97 citations), demonstrates how evolutionary algorithms can simultaneously optimize both where a vision system looks and what features it extracts, mimicking biological visual attention. His earlier foundational study, "An Evolutionary Active-Vision System" (2002, 22 citations), introduced a groundbreaking artificial retina controlled by an evolutionary recurrent neural network, capable of autonomously zooming, panning, and adjusting filtering strategies to discriminate shapes without explicit programming. This work established a framework for self-adaptive visual processing that has influenced subsequent research in embodied cognition, autonomous robotics, and biologically inspired computer vision. Kato's research elegantly bridges evolutionary computation and active perception, showing that intelligent vision emerges not from static algorithms but from the dynamic interplay between movement, attention, and learning.