FMix: Enhancing Mixed Sample Data Augmentation

By Ethan Harris, Antonia Marcu, Matthew Painter, Mahesan Niranjan, Adam Prügel-Bennett, Jonathon S. Hare. 2020

183 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Sharpness-Aware Minimization for Efficiently Improving Generalization (2020, background): “…for modifying the training process have been proposed, including dropout (Srivastava et al., 2014), batch normalization (Ioffe & Szegedy, 2015), stochastic depth (Huang et al., 2016), data augmentation (Cubuk et al., 2018), and mixed sample augmentations (Zhang et al., 2017; Harris et al., 2020).”
Milking CowMask for Semi-Supervised Image Classification (2020, methodology): “We note that the concurrent work FMix (Harris et al., 2020) uses an inverse Fourier transform to generate masks with a similar visual appearance.”
Review of Image Classification Algorithms Based on Convolutional Neural Networks (2021, background): “Mixed Sample Data Augmentation (MSDA) [245,246,251] is to randomly mix two training samples and their labels according to a certain ratio, which can not only reduce the misidentification of some difficult samples but also improve the robustness of the model and make it more stable during training.”
Image Data Augmentation for Deep Learning: A Survey (2022, background): “[Harris et al., 2020] proposes Fmix that uses random binary masks obtained by applying a threshold to low-frequency images sampled from Fourier space.”
i-Mix: A Domain-Agnostic Strategy for Contrastive Representation Learning (2021, background): “MixUp would not be directly applicable to some domains, such as point clouds, but its adaptation can be effective (Harris et al., 2020). i -Mix is a kind of data augmentation for better generalization in contrastive representation learning, resulting in better performances on downstream tasks.”
Decoupled Mixup for Data-efficient Learning (2022, unknown): “For example, the mixed samples in early hand-crafted mixing policies [42,41,11] (referred to as static methods) might suffer from over-smoothing when mixed samples are semantic inconsistency with the mixed labels, as shown in Figure 1.”
ResizeMix: Mixing Data with Preserved Object Information and True Labels (2020, methodology): “Somewhat differently, FMix [22] sets a threshold for the low-frequency parts in the image to get the saliency masks for mixing images.”
Improving Augmentation and Evaluation Schemes for Semantic Image Synthesis (2020, background): “Recently, label-perturbing data augmentation methods have received a lot of attention [70, 60, 67, 17], proposing to create augmented images by mixing samples from different classes and interpolating their labels accordingly.”