Data augmentation is a ubiquitous technique used to provide robustness t...
Despite the fast progress in training specialized models for various tas...
Novel computer vision architectures monopolize the spotlight, but the im...
Building instance segmentation models that are data-efficient and can ha...
Materials design enables technologies critical to humanity, including
co...
Including prior knowledge is important for effective machine learning mo...
Pre-training is a dominant paradigm in computer vision. For example,
sup...
Supervised learning in large discriminative models is a mainstay for mod...
Supervised learning in large discriminative models is a mainstay for mod...
Though data augmentation has become a standard component of deep neural
...
Semi-supervised learning (SSL) provides an effective means of leveraging...
A large fraction of computational science involves simulating the dynami...
Modern deep neural networks can achieve high accuracy when the training
...
We improve the recently-proposed "MixMatch" semi-supervised learning
alg...
Recent work has shown that data augmentation has the potential to
signif...
Data augmentation is a critical component of training deep learning mode...
Achieving robustness to distributional shift is a longstanding and
chall...
State-of-the art vision models can achieve superhuman performance on ima...
Deploying machine learning systems in the real world requires both high
...
We present SpecAugment, a simple data augmentation method for speech
rec...
In this paper, we take a closer look at data augmentation for images, an...
Semi-supervised learning (SSL) provides a powerful framework for leverag...
It is becoming increasingly clear that many machine learning classifiers...