ImageNet-trained CNNs are biased towards texture (2019)

Re:ImageNet-trained CNNs are biased towards texture (2019) by wodenokoto on Monday April 20, 2020 @03:59AM (#22922041)

My favorite titled ml article [1], from 2015, seems to argue for the same conclusion - that it is textures not shapes that drives recognition in CNN

[1] “Suddenly, a leopard print sofa appears”, http://rocknrollnerd.github.io/ml/2015/05/27/leopard-sofa.ht...

[ Parent | on HN ]

Re:ImageNet-trained CNNs are biased towards texture (2019) by modeless on Monday April 20, 2020 @04:15AM (#22922102)

> it is textures not shapes that drives recognition in CNN

This is a misinterpretation of the paper. It's not CNNs that are biased toward textures, it's the ImageNet dataset. As they prove by training a CNN biased toward shapes, using a modified dataset.

[ Parent | on HN ]

Re:ImageNet-trained CNNs are biased towards texture (2019) by fxtentacle on Monday April 20, 2020 @05:03AM (#22922275)

I thought the same. Captain Obvious called a second time. Of course CNNs are still heavily biased towards texture, because nothing has fundamentally changed in the past 5 years.

[ Parent | on HN ]