Fixing the train-test resolution discrepancy: FixEfficientNet
share
This note complements the paper "Fixing the train-test resolution discrepancy" that introduced the FixRes method. First, we show that this strategy is advantageously combined with recent training recipes from the literature. Most importantly, we provide new results for the EfficientNet architecture. The resulting network, called FixEfficientNet, significantly outperforms the initial architecture with the same number of parameters. For instance, our FixEfficientNet-B0 trained without additional training data achieves 79.3 absolute improvement over the Noisy student EfficientNet-B0 trained with 300M unlabeled images and +1.7 adversarial examples. An EfficientNet-L2 pre-trained with weak supervision on 300M unlabeled images and further optimized with FixRes achieves 88.5 accuracy (top-5: 98.7 ImageNet with a single crop.
READ FULL TEXT
