# Revisiting Distillation and Incremental Classifier Learning

arXiv: https://arxiv.org/abs/1807.02802 · HTML: https://mlsanity.com/paper/1807.02802/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Jul 2018.

## Built on by (most cited first, 19 of 19)

- [Online Continual Learning in Image Classification: An Empirical Survey](https://mlsanity.com/paper/2101.10423/) (2021)
- [Continual Normalization: Rethinking Batch Normalization for Online Continual Learning](https://mlsanity.com/paper/2203.16102/) (2022)
- [DualNet: Continual Learning, Fast and Slow](https://mlsanity.com/paper/2110.00175/) (2021)
- [Prototype-Sample Relation Distillation: Towards Replay-Free Continual Learning](https://mlsanity.com/paper/2303.14771/) (2023)
- [Generalized Knowledge Distillation via Relationship Matching](https://mlsanity.com/paper/2205.01915/) (2022)
- [Subspace Distillation for Continual Learning](https://mlsanity.com/paper/2307.16419/) (2023)
- [TKIL: Tangent Kernel Approach for Class Balanced Incremental Learning](https://mlsanity.com/paper/2206.08492/) (2022)
- [Queried Unlabeled Data Improves and Robustifies Class-Incremental Learning](https://mlsanity.com/paper/2206.07842/) (2022)
- [Robustness-preserving Lifelong Learning via Dataset Condensation](https://mlsanity.com/paper/2303.04183/) (2023)
- [Multi-Teacher Knowledge Distillation for Incremental Implicitly-Refined Classification](https://mlsanity.com/paper/2202.11384/) (2022)
- [PlaStIL: Plastic and Stable Memory-Free Class-Incremental Learning](https://mlsanity.com/paper/2209.06606/) (2022)
- [A Comparative Study of Calibration Methods for Imbalanced Class Incremental Learning](https://mlsanity.com/paper/2202.00386/) (2022)
- [APP: Anytime Progressive Pruning](https://mlsanity.com/paper/2204.01640/) (2022)
- [PBES: PCA Based Exemplar Sampling Algorithm for Continual Learning](https://mlsanity.com/paper/2312.09352/) (2023)
- [PTMs-TSCIL Pre-Trained Models Based Class-Incremental Learning](https://mlsanity.com/paper/2503.07153/) (2025)
- [Continual Learning, Fast and Slow](https://mlsanity.com/paper/2209.02370/) (2022)
- [Single image calibration using knowledge distillation approaches](https://mlsanity.com/paper/2212.02379/) (2022)
- [Detachedly Learn a Classifier for Class-Incremental Learning](https://mlsanity.com/paper/2302.11730/) (2023)
- [MEGG: Replay via Maximally Extreme GGscore in Incremental Learning for Neural Recommendation Models](https://mlsanity.com/paper/2509.07319/) (2025)

## Abstract (arXiv)

One of the key differences between the learning mechanism of humans and Artificial Neural Networks (ANNs) is the ability of humans to learn one task at a time. ANNs, on the other hand, can only learn multiple tasks simultaneously. Any attempts at learning new tasks incrementally cause them to completely forget about previous tasks. This lack of ability to learn incrementally, called Catastrophic Forgetting, is considered a major hurdle in building a true AI system. In this paper, our goal is to isolate the truly effective existing ideas for incremental learning from those that only work under certain conditions. To this end, we first thoroughly analyze the current state of the art (iCaRL) method for incremental learning and demonstrate that the good performance of the system is not because of the reasons presented in the existing literature. We conclude that the success of iCaRL is primarily due to knowledge distillation and recognize a key limitation of knowledge distillation, i.e, it often leads to bias in classifiers. Finally, we propose a dynamic threshold moving algorithm that is able to successfully remove this bias. We demonstrate the effectiveness of our algorithm on CIFAR100 and MNIST datasets showing near-optimal results. Our implementation is available at https://github.com/Khurramjaved96/incremental-learning.