Drawing Multiple Augmentation Samples Per Image During Training Efficiently Decreases Test Error
By Stanislav Fort, Andrew Brock, Razvan Pascanu, Soham De, Samuel L. Smith. 2021
36 citations recorded by the source corpus.
In computer vision, it is standard practice to draw a single sample from the data augmentation procedure for each unique image in the mini-batch. However recent work has suggested drawing multiple samples can achieve higher test accuracies. In this work, we provide a detailed empirical evaluation of how the number of augmentation samples per unique image influences model performance on held out data when training deep ResNets. We demonstrate drawing multiple samples per image consistently enhances the test accuracy achieved for both small and large batch training. Crucially, this benefit arises even if different numbers of augmentations per image perform the same number of parameter updates and gradient evaluations (requiring the same total compute). Although prior work has found variance in the gradient estimate arising from subsampling the dataset has an implicit regularization benefit, our experiments suggest variance which arises from the data augmentation process harms generalization. We apply these insights to the highly performant NFNet-F5, achieving 86.8$\%$ top-1 w/o extra data on ImageNet.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging, Optimization & Training
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Soham De
- Jamie Hayes
- Borja Balle
- Samuel L. Smith
- Leonard Berrada
- Connor Shorten
- Taghi M. Khoshgoftaar
- Borko Furht
In Computer Vision and Pattern Recognition
- Soham De
- Jamie Hayes
- Samuel L. Smith
- Leonard Berrada
- Borja Balle
- Connor Shorten
- Taghi M. Khoshgoftaar
- Borko Furht
In Optimization & Training
- Soham De
- Samuel L. Smith
- Jamie Hayes
- Leonard Berrada
- Borja Balle
- Taghi M. Khoshgoftaar
- Andrew Brock
- Quoc V. Le
Citations in the corpus
- Text Data Augmentation for Deep Learning (2021, background): “Fort et al. [128] and Hoffer et al. [129] illustrate how increasing augmentation multiplicity can improve performance.”
- Unlocking High-Accuracy Differentially Private Image Classification through Scale (2022, methodology): “, 2019), a modification to DP-SGD which we call augmentation multiplicity (Fort et al., 2021; Hoffer et al., 2019; Touvron et al., 2021a) and parameter averaging (Polyak and Juditsky, 1992).”