Calibrating Class Activation Maps for Long-Tailed Visual Recognition

TL;DR

This paper introduces a novel calibration module and normalization technique to improve long-tailed visual recognition, significantly enhancing performance on multiple benchmarks by addressing class imbalance issues.

Contribution

It proposes a Class Activation Map Calibration (CAMC) module and normalized classifiers, offering new methods to mitigate bias towards head classes in long-tailed distributions.

Findings

Achieved state-of-the-art results on five benchmarks.

Improved tail class accuracy without sacrificing head class performance.

Validated effectiveness of calibration and normalization techniques.

Abstract

Real-world visual recognition problems often exhibit long-tailed distributions, where the amount of data for learning in different categories shows significant imbalance. Standard classification models learned on such data distribution often make biased predictions towards the head classes while generalizing poorly to the tail classes. In this paper, we present two effective modifications of CNNs to improve network learning from long-tailed distribution. First, we present a Class Activation Map Calibration (CAMC) module to improve the learning and prediction of network classifiers, by enforcing network prediction based on important image regions. The proposed CAMC module highlights the correlated image regions across data and reinforces the representations in these areas to obtain a better global representation for classification. Furthermore, we investigate the use of normalized classifiers for representation learning in long-tailed problems. Our empirical study demonstrates that by simply scaling the outputs of the classifier with an appropriate scalar, we can effectively improve the classification accuracy on tail classes without losing the accuracy of head classes. We conduct extensive experiments to validate the effectiveness of our design and we set new state-of-the-art performance on five benchmarks, including ImageNet-LT, Places-LT, iNaturalist 2018, CIFAR10-LT, and CIFAR100-LT.