RESEARCH

COMPUTER VISION

Fixing the train-test resolution discrepancy

December 09, 2019

Abstract

Data-augmentation is key to the training of neural networks for image classifi- cation. 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 128×128 images, and 79.8% with one trained at 224×224. A ResNeXt-101 32x48d pre-trained with weak supervision on 940 million 224×224 images and further optimized with our technique for test resolution 320×320 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.

Download the Paper

AUTHORS

Written by

Andrea Vedaldi

Hervé Jegou

Hugo Touvron

Matthijs Douze

Publisher

NeurIPS

Research Topics

Computer Vision

Related Publications

December 13, 2022

NLP

COMPUTER VISION

Efficient Self-supervised Learning with Contextualized Target Representations for Vision, Speech and Language

Michael Auli, Alexei Baevski, Arun Babu, Wei-Ning Hsu

December 13, 2022

November 28, 2022

RESEARCH

CORE MACHINE LEARNING

Neural Attentive Circuits

Nicolas Ballas, Bernhard Schölkopf, Chris Pal, Francesco Locatello, Li Erran, Martin Weiss, Nasim Rahaman, Yoshua Bengio

November 28, 2022

November 27, 2022

RESEARCH

Near Instance-Optimal PAC Reinforcement Learning for Deterministic MDPs

Andrea Tirinzoni, Aymen Al Marjani, Emilie Kaufmann

November 27, 2022

November 16, 2022

RESEARCH

NLP

Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Models

Kushal Tirumala, Aram H. Markosyan, Armen Aghajanyan, Luke Zettlemoyer

November 16, 2022

Help Us Pioneer The Future of AI

We share our open source frameworks, tools, libraries, and models for everything from research exploration to large-scale production deployment.