Timezone: »

Fixing the train-test resolution discrepancy
Hugo Touvron · Andrea Vedaldi · Matthijs Douze · Herve Jegou

Thu Dec 12 10:45 AM -- 12:45 PM (PST) @ East Exhibition Hall B + C #141

Data-augmentation is key to the training of neural networks for image classification. This paper first shows that existing augmentations induce a significant discrepancy between the size of the objects seen by the classifier at train and test time: in fact, a lower train resolution improves the classification at test time!

We then propose a simple strategy to optimize the classifier performance, that employs different train and test resolutions. It relies on a computationally cheap fine-tuning of the network at the test resolution. This enables training strong classifiers using small training images, and therefore significantly reduce the training time. For instance, we obtain 77.1% top-1 accuracy on ImageNet with a ResNet-50 trained on 128x128 images, and 79.8% with one trained at 224x224.

A ResNeXt-101 32x48d pre-trained with weak supervision on 940 million 224x224 images and further optimized with our technique for test resolution 320x320 achieves 86.4% top-1 accuracy (top-5: 98.0%). To the best of our knowledge this is the highest ImageNet single-crop accuracy to date.

Author Information

Hugo Touvron (Facebook AI Research)
Andrea Vedaldi (Facebook AI Research and University of Oxford)
Matthijs Douze (Facebook AI Research)
Herve Jegou (Facebook AI Research)

More from the Same Authors