ICML 2019

EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

ICML 2019 tensorflow/tpu

Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available.

 SOTA for Image Classification on Stanford Cars (using extra training data)

IMAGE CLASSIFICATION NEURAL ARCHITECTURE SEARCH TRANSFER LEARNING