# Distilling Balanced Knowledge from a Biased Teacher

arXiv: https://arxiv.org/abs/2506.18496 · HTML: https://mlsanity.com/paper/2506.18496/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in General knowledge distillation methods, one of 12 directions in Knowledge distillation.
- Builds on 45 of its 53 references that are in the map.
- First posted on arXiv 23 Jun 2025.

## Placed in

- Knowledge distillation / [General knowledge distillation methods](https://mlsanity.com/distill/general-knowledge-distillation-methods/)

## Builds on (most cited first, 30 of 45)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [Wide Residual Networks](https://mlsanity.com/paper/1605.07146/) (2016)
- [FitNets: Hints for Thin Deep Nets](https://mlsanity.com/paper/1412.6550/) (2014)
- [Knowledge Distillation: A Survey](https://mlsanity.com/paper/2006.05525/) (2020)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices](https://mlsanity.com/paper/1707.01083/) (2017)
- [ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design](https://mlsanity.com/paper/1807.11164/) (2018)
- [Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer](https://mlsanity.com/paper/1612.03928/) (2016)
- [Relational Knowledge Distillation](https://mlsanity.com/paper/1904.05068/) (2019)
- [Contrastive Representation Distillation](https://mlsanity.com/paper/1910.10699/) (2019)
- [Large-Scale Long-Tailed Recognition in an Open World](https://mlsanity.com/paper/1904.05160/) (2019)
- [Deep Mutual Learning](https://mlsanity.com/paper/1706.00384/) (2017)
- [Improved Knowledge Distillation via Teacher Assistant](https://mlsanity.com/paper/1902.03393/) (2019)
- [Patient Knowledge Distillation for BERT Model Compression](https://mlsanity.com/paper/1908.09355/) (2019)
- [Deep Long-Tailed Learning: A Survey](https://mlsanity.com/paper/2110.04596/) (2021)
- [Decoupled Knowledge Distillation](https://mlsanity.com/paper/2203.08679/) (2022)
- [Distilling Knowledge via Knowledge Review](https://mlsanity.com/paper/2104.09044/) (2021)
- [A Survey on Knowledge Distillation of Large Language Models](https://mlsanity.com/paper/2402.13116/) (2024)
- [Knowledge Distillation from A Stronger Teacher](https://mlsanity.com/paper/2205.10536/) (2022)
- [Learning From Multiple Experts: Self-paced Knowledge Distillation for Long-tailed Classification](https://mlsanity.com/paper/2001.01536/) (2020)
- [Cross-Image Relational Knowledge Distillation for Semantic Segmentation](https://mlsanity.com/paper/2204.06986/) (2022)
- [Long-Tailed Recognition via Weight Balancing](https://mlsanity.com/paper/2203.14197/) (2022)
- [A Survey on Long-Tailed Visual Recognition](https://mlsanity.com/paper/2205.13775/) (2022)
- [Self Supervision to Distillation for Long-Tailed Visual Recognition](https://mlsanity.com/paper/2109.04075/) (2021)
- [Distilling Virtual Examples for Long-tailed Recognition](https://mlsanity.com/paper/2103.15042/) (2021)
- [Class Attention Transfer Based Knowledge Distillation](https://mlsanity.com/paper/2304.12777/) (2023)

## Abstract (arXiv)

Conventional knowledge distillation, designed for model compression, fails on long-tailed distributions because the teacher model tends to be biased toward head classes and provides limited supervision for tail classes. We propose Long-Tailed Knowledge Distillation (LTKD), a novel framework that reformulates the conventional objective into two components: a cross-group loss, capturing mismatches in prediction distributions across class groups (head, medium, and tail), and a within-group loss, capturing discrepancies within each group's distribution. This decomposition reveals the specific sources of the teacher's bias. To mitigate the inherited bias, LTKD introduces (1) a rebalanced cross-group loss that calibrates the teacher's group-level predictions and (2) a reweighted within-group loss that ensures equal contribution from all groups. Extensive experiments on CIFAR-100-LT, TinyImageNet-LT, and ImageNet-LT demonstrate that LTKD significantly outperforms existing methods in both overall and tail-class accuracy, thereby showing its ability to distill balanced knowledge from a biased teacher for real-world applications.