# Meta Feature Modulator for Long-tailed Recognition

arXiv: https://arxiv.org/abs/2008.03428 · HTML: https://mlsanity.com/paper/2008.03428/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Aug 2020.

## Built on by (most cited first, 7 of 7)

- [A Survey on Long-Tailed Visual Recognition](https://mlsanity.com/paper/2205.13775/) (2022)
- [Fill-Up: Balancing Long-Tailed Data with Generative Models](https://mlsanity.com/paper/2306.07200/) (2023)
- [CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning](https://mlsanity.com/paper/2202.05613/) (2022)
- [Imbalanced Semi-supervised Learning with Bias Adaptive Classifier](https://mlsanity.com/paper/2207.13856/) (2022)
- [CBA: Improving Online Continual Learning via Continual Bias Adaptor](https://mlsanity.com/paper/2308.06925/) (2023)
- [Learning an Explicit Hyperparameter Prediction Function Conditioned on Tasks](https://mlsanity.com/paper/2107.02378/) (2021)
- [Dual-CBA: Improving Online Continual Learning via Dual Continual Bias Adaptors from a Bi-level Optimization Perspective](https://mlsanity.com/paper/2408.13991/) (2024)

## Abstract (arXiv)

Deep neural networks often degrade significantly when training data suffer from class imbalance problems. Existing approaches, e.g., re-sampling and re-weighting, commonly address this issue by rearranging the label distribution of training data to train the networks fitting well to the implicit balanced label distribution. However, most of them hinder the representative ability of learned features due to insufficient use of intra/inter-sample information of training data. To address this issue, we propose meta feature modulator (MFM), a meta-learning framework to model the difference between the long-tailed training data and the balanced meta data from the perspective of representation learning. Concretely, we employ learnable hyper-parameters (dubbed modulation parameters) to adaptively scale and shift the intermediate features of classification networks, and the modulation parameters are optimized together with the classification network parameters guided by a small amount of balanced meta data. We further design a modulator network to guide the generation of the modulation parameters, and such a meta-learner can be readily adapted to train the classification network on other long-tailed datasets. Extensive experiments on benchmark vision datasets substantiate the superiority of our approach on long-tailed recognition tasks beyond other state-of-the-art methods.