Configural relations in humans and deep convolutional neural networks

HIGHLIGHTS

SUMMARY

Or surface feature segmentation (Shi and Malik, 2000; Shotton et_al, 2009). Deep networks trained for object recognition also appear to predict human behavior in judging the similarity between objects (Peterson et_al, 2016), the memorability of objects (Dubey et_al, 2015), and the saliency of regions in an image (Kümmerer et_al, 2014). Relevant relations as stimuli for vision often involve considerable complexity (Johansson, 1978; Gibson, 1979; Ullman, 1979; Marr, 1982; Palmer et_al, 2006; Baker and Kellman, 2018). Evidence indicates the abstract nature of these and other perceptual representations (e_g, Izard et_al, 2009; Hummel, 2011; Baker . . .

 



If you want to have access to all the content you need to log in!

Thanks :)

Username or Email

Password

Remember me

Lost your password?

If you don't have an account, you can create one here.