Feature Space Augmentation for Long-Tailed Data

By Peng Chu, Xiao Bian, Shaopeng Liu, Haibin Ling. 2020

289 citations recorded by the source corpus.

Real-world data often follow a long-tailed distribution as the frequency of each class is typically different. For example, a dataset can have a large number of under-represented classes and a few classes with more than sufficient data. However, a model to represent the dataset is usually expected to have reasonably homogeneous performances across classes. Introducing class-balanced loss and advanced methods on data re-sampling and augmentation are among the best practices to alleviate the data imbalance problem. However, the other part of the problem about the under-represented classes will have to rely on additional knowledge to recover the missing information. In this work, we present a novel approach to address the long-tailed problem by augmenting the under-represented classes in the feature space with the features learned from the classes with ample samples. In particular, we decompose the features of each class into a class-generic component and a class-specific component using class activation maps. Novel samples of under-represented classes are then generated on the fly during training stages by fusing the class-specific features from the under-represented classes with the class-generic features from confusing classes. Our results on different datasets such as iNaturalist, ImageNet-LT, Places-LT and a long-tailed version of CIFAR have shown the state of the art performances.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging, Representation & Self-Supervised Learning

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

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

In Representation & Self-Supervised Learning

Citations in the corpus

Grafit: Learning fine-grained image representations with coarse labels (2020, positive): “7 iNaturalist 2018 ResNet-152 [12] 60M 224 69.”
Long-Tailed Class Incremental Learning (2022, methodology): “Two-stage methods have shown state-of-the-art performance in long-tailed recognition [8,16,47,50].”
Long-tailed Recognition by Routing Diverse Distribution-Aware Experts (2020, unknown): “For example, Liu et al. (2020), Chu et al. (2020), and Kim et al. (2020) generate augmented samples in the feature space to supplement tail class samples.”
Contrastive Learning based Hybrid Networks for Long-Tailed Image Classification (2021, methodology): “Typical methods include data re-sampling [1, 8, 35, 4, 28], loss re-weighting [26, 7, 30, 33], margin modification [3], and data augmentation [19, 6, 29].”
Distribution Alignment: A Unified Framework for Long-tail Visual Recognition (2021, background): “Other work aim to enhance the representation or classifier head of tail categories by transferring knowledge from the head classes [41, 40, 26, 46, 8, 43, 42].”