# Reducing Class-wise Confusion for Incremental Learning with Disentangled Manifolds

arXiv: https://arxiv.org/abs/2503.17677 · HTML: https://mlsanity.com/paper/2503.17677/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 40 references that are in the map.
- First posted on arXiv 22 Mar 2025.

## Builds on (most cited first, 21 of 21)

- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [iCaRL: Incremental Classifier and Representation Learning](https://mlsanity.com/paper/1611.07725/) (2016)
- [DER: Dynamically Expandable Representation for Class Incremental Learning](https://mlsanity.com/paper/2103.16788/) (2021)
- [Semantic Drift Compensation for Class-Incremental Learning](https://mlsanity.com/paper/2004.00440/) (2020)
- [FOSTER: Feature Boosting and Compression for Class-Incremental Learning](https://mlsanity.com/paper/2204.04662/) (2022)
- [RanPAC: Random Projections and Pre-trained Models for Continual Learning](https://mlsanity.com/paper/2307.02251/) (2023)
- [A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning](https://mlsanity.com/paper/2205.13218/) (2022)
- [Semantic Autoencoder for Zero-Shot Learning](https://mlsanity.com/paper/1704.08345/) (2017)
- [Expandable Subspace Ensemble for Pre-Trained Model-Based Class-Incremental Learning](https://mlsanity.com/paper/2403.12030/) (2024)
- [FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning](https://mlsanity.com/paper/2309.14062/) (2023)
- [PyCIL: A Python Toolbox for Class-Incremental Learning](https://mlsanity.com/paper/2112.12533/) (2021)
- [Learning Representation for Clustering via Prototype Scattering and Positive Sampling](https://mlsanity.com/paper/2111.11821/) (2021)
- [Divide and not forget: Ensemble of selectively trained experts in Continual Learning](https://mlsanity.com/paper/2401.10191/) (2024)
- [Error Sensitivity Modulation based Experience Replay: Mitigating Abrupt Representation Drift in Continual Learning](https://mlsanity.com/paper/2302.11344/) (2023)
- [Variational Prototyping-Encoder: One-Shot Learning with Prototypical Images](https://mlsanity.com/paper/1904.08482/) (2019)
- [Learning Manifold Dimensions with Conditional Variational Autoencoders](https://mlsanity.com/paper/2302.11756/) (2023)
- [Class Incremental Learning via Likelihood Ratio Based Task Prediction](https://mlsanity.com/paper/2309.15048/) (2023)
- [Exploring Data Geometry for Continual Learning](https://mlsanity.com/paper/2304.03931/) (2023)
- [Multi-Granularity Regularized Re-Balancing for Class Incremental Learning](https://mlsanity.com/paper/2206.15189/) (2022)
- [Dynamic Sub-graph Distillation for Robust Semi-supervised Continual Learning](https://mlsanity.com/paper/2312.16409/) (2023)

## Built on by (most cited first, 4 of 4)

- [Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging](https://mlsanity.com/paper/2605.18608/) (2026)
- [C-Flat++: Towards a More Efficient and Powerful Framework for Continual Learning](https://mlsanity.com/paper/2508.18860/) (2025)
- [CIFNet: An Analytic Neural Learning Framework for Efficient and Calibrated Class-Incremental Learning](https://mlsanity.com/paper/2509.11285/) (2025)
- [Grow, Assess, Compress: Adaptive Backbone Scaling for Memory-Efficient Class Incremental Learning](https://mlsanity.com/paper/2603.08426/) (2026)

## Abstract (arXiv)

Class incremental learning (CIL) aims to enable models to continuously learn new classes without catastrophically forgetting old ones. A promising direction is to learn and use prototypes of classes during incremental updates. Despite simplicity and intuition, we find that such methods suffer from inadequate representation capability and unsatisfied feature overlap. These two factors cause class-wise confusion and limited performance. In this paper, we develop a Confusion-REduced AuTo-Encoder classifier (CREATE) for CIL. Specifically, our method employs a lightweight auto-encoder module to learn compact manifold for each class in the latent subspace, constraining samples to be well reconstructed only on the semantically correct auto-encoder. Thus, the representation stability and capability of class distributions are enhanced, alleviating the potential class-wise confusion problem. To further distinguish the overlapped features, we propose a confusion-aware latent space separation loss that ensures samples are closely distributed in their corresponding low-dimensional manifold while keeping away from the distributions of features from other classes. Our method demonstrates stronger representational capacity and discrimination ability by learning disentangled manifolds and reduces class confusion. Extensive experiments on multiple datasets and settings show that CREATE outperforms other state-of-the-art methods up to 5.41%.