摘要: This paper does not describe a novel method. Instead, it studies a straightforward, incremental, yet must-know baseline given the recent progress in computer vision: self …
Aapo Kyrola, Piotr Dollár, Lukasz Wesolowski, Yangqing Jia, Andrew Tulloch, Kaiming He, Ross B. Girshick, Priya Goyal, Pieter Noordhuis, Accurate, Large Minibatch SGD: Training ImageNet in 1 HourarXiv: Computer Vision and Pattern Recognition. ,(2017)