2011 Participants
Paik Chow
The role of colour in object recognition by humans is controversial. Most studies conclude that shape is the primary cue for recognition, while surface properties such as colour and texture play incidental roles. Recent results suggest, though, that colour is faster and more accurate than either texture or shape in driving object classification. Here we investigate the relative advantages for colour and texture in two behavioural tasks using natural images: visual search and material categorisation. The results will be important for understanding human visual cognition as well as guiding computer vision algorithms that perform rapid recognition tasks in human-like fashion.
Funding Source: Newcastle University