이번 포스팅은 논문 ImageNet - Train CNNs are biased towards texture; increasing shape bias improves accuracy and robustness을 읽고 읽은 내용을 포스팅한다.
Continue reading
![](http://ssaru.github.io/2019/08/06/20190806-How_the_Eye_of_AI_Sees/cover.jpeg)
Software Engineer in the field of Machine Learning
Data scientists in Viva Republic