Avoiding overfitting: A survey on regularization methods for convolutional neural networks

CFGD Santos, JP Papa - ACM Computing Surveys (CSUR), 2022 - dl.acm.org
Several image processing tasks, such as image classification and object detection, have
been significantly improved using Convolutional Neural Networks (CNN). Like ResNet and …

A survey of mix-based data augmentation: Taxonomy, methods, applications, and explainability

C Cao, F Zhou, Y Dai, J Wang, K Zhang - ACM Computing Surveys, 2022 - dl.acm.org
Data augmentation (DA) is indispensable in modern machine learning and deep neural
networks. The basic idea of DA is to construct new training data to improve the model's …

MedViT: a robust vision transformer for generalized medical image classification

ON Manzari, H Ahmadabadi, H Kashiani… - Computers in Biology …, 2023 - Elsevier
Abstract Convolutional Neural Networks (CNNs) have advanced existing medical systems
for automatic disease diagnosis. However, there are still concerns about the reliability of …

How does mixup help with robustness and generalization?

L Zhang, Z Deng, K Kawaguchi, A Ghorbani… - arXiv preprint arXiv …, 2020 - arxiv.org
Mixup is a popular data augmentation technique based on taking convex combinations of
pairs of examples and their labels. This simple technique has been shown to substantially …

Uncertainty baselines: Benchmarks for uncertainty & robustness in deep learning

Z Nado, N Band, M Collier, J Djolonga… - arXiv preprint arXiv …, 2021 - arxiv.org
High-quality estimates of uncertainty and robustness are crucial for numerous real-world
applications, especially for deep learning which underlies many deployed ML systems. The …

Fmix: Enhancing mixed sample data augmentation

E Harris, A Marcu, M Painter, M Niranjan… - arXiv preprint arXiv …, 2020 - arxiv.org
Mixed Sample Data Augmentation (MSDA) has received increasing attention in recent
years, with many successful variants such as MixUp and CutMix. By studying the mutual …

A unified analysis of mixed sample data augmentation: A loss function perspective

C Park, S Yun, S Chun - Advances in Neural Information …, 2022 - proceedings.neurips.cc
We propose the first unified theoretical analysis of mixed sample data augmentation
(MSDA), such as Mixup and CutMix. Our theoretical results show that regardless of the …

Mixmo: Mixing multiple inputs for multiple outputs via deep subnetworks

A Ramé, R Sun, M Cord - Proceedings of the IEEE/CVF …, 2021 - openaccess.thecvf.com
Recent strategies achieved ensembling"" for free"" by fitting concurrently diverse
subnetworks inside a single base network. The main idea during training is that each …

Survey: Image mixing and deleting for data augmentation

H Naveed, S Anwar, M Hayat, K Javed… - Engineering Applications of …, 2024 - Elsevier
Neural networks are prone to overfitting and memorizing data patterns. To avoid over-fitting
and enhance their generalization and performance, various methods have been suggested …

The benefits of mixup for feature learning

D Zou, Y Cao, Y Li, Q Gu - International Conference on …, 2023 - proceedings.mlr.press
Mixup, a simple data augmentation method that randomly mixes two data points via linear
interpolation, has been extensively applied in various deep learning applications to gain …