Class Incremental Learning

Class incremental learning (CIL) focuses on training machine learning models to continuously learn new classes of data without forgetting previously learned ones, a crucial challenge for real-world applications with evolving data streams. Current research emphasizes techniques like dynamic model architectures (e.g., adding task-specific adapters), generative replay methods to synthesize past data, and the use of pre-trained models to leverage existing knowledge. These advancements aim to improve accuracy and fairness while addressing issues like catastrophic forgetting and data imbalance, impacting fields such as medical image analysis, sound source localization, and personalized AI systems.

Papers

January 3, 2024

Learning Prompt with Distribution-Based Feature Replay for Few-Shot Class-Incremental Learning
Zitong Huang, Ze Chen, Zhixing Chen, Erjin Zhou, Xinxing Xu, Rick Siow Mong Goh, Yong Liu, Wangmeng Zuo, Chunmei Feng
Vision Language Model Class Incremental Learning Training Image Latent Replay Mini ImageNet

January 1, 2024

Federated Class-Incremental Learning with New-Class Augmented Self-Distillation
Zhiyuan Wu, Tianliu He, Sheng Sun, Yuwei Wang, Min Liu, Bo Gao, Xuefeng Jiang
Class Incremental Learning Self Distillation Federated Prompt Cooperation

December 20, 2023

Fine-Grained Knowledge Selection and Restoration for Non-Exemplar Class Incremental Learning
Jiang-Tian Zhai, Xialei Liu, Lu Yu, Ming-Ming Cheng
Fine Grained Catastrophic Forgetting Incremental Learning Class Incremental Learning Restoration Quality Fine Grained Knowledge

December 14, 2023

December 8, 2023

Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration
Qi-Wei Wang, Da-Wei Zhou, Yi-Kai Zhang, De-Chuan Zhan, Han-Jia Ye
Class Incremental Learning Unseen Class General Class Prototypical Calibration

December 7, 2023

TLCE: Transfer-Learning Based Classifier Ensembles for Few-Shot Class-Incremental Learning
Shuangmei Wang, Yang Cao, Tieru Wu
Training Data Pre Trained Class Incremental Learning Ensemble Classifier Shot Class Incremental Learning Episodic Training

December 6, 2023

Class Incremental Learning for Adversarial Robustness
Seungju Cho, Hongsin Lee, Changick Kim
Adversarial Example Adversarial Training Adversarial Robustness Incremental Learning Class Incremental Learning

December 5, 2023

MIND: Multi-Task Incremental Network Distillation
Jacopo Bonato, Francesco Pelosin, Luigi Sabetta, Alessandro Nicolosi
Class Incremental Learning Human Mind Domain Incremental Learning Replay Strategy Data Replay Rehearsal Free Class Incremental Learning

November 30, 2023

Prompt-Based Exemplar Super-Compression and Regeneration for Class-Incremental Learning
Ruxiao Duan, Yaoyao Liu, Jieneng Chen, Adam Kortylewski, Alan Yuille
Class Incremental Learning Pre Trained Diffusion Model Exemplar Image Diffusion Based Data Augmentation

November 22, 2023

Density Distribution-based Learning Framework for Addressing Online Continual Learning Challenges
Shilin Zhang, Jiahui Wang
Continual Learning Class Incremental Learning Online Continual Learning Distribution Learning Classical Machine Learning Generative Factor

November 19, 2023

From Categories to Classifiers: Name-Only Continual Learning by Exploring the Web
Ameya Prabhu, Hasan Abed Al Kader Hammoud, Ser-Nam Lim, Bernard Ghanem, Philip H.S. Torr, Adel Bibi
Continual LEArning Continual Learning Class Incremental Learning Simple Classifier Category Shift

November 15, 2023

Combining Past, Present and Future: A Self-Supervised Approach for Class Incremental Learning
Xiaoshuang Chen, Zhongyi Sun, Ke Yan, Shouhong Ding, Hongtao Lu
Self Supervised Incremental Learning Class Incremental Learning Speech Presence Future Reasoning Self Supervised Approach Novel Class

October 31, 2023

October 30, 2023

Look At Me, No Replay! SurpriseNet: Anomaly Detection Inspired Class Incremental Learning
Anton Lee, Yaqian Zhang, Heitor Murilo Gomes, Albert Bifet, Bernhard Pfahringer
Continual LEArning Class Incremental Learning RePLAy Loss Incremental Training Surprise Bound Cross Task Knowledge Knowledge Update DMIS

October 24, 2023

Wakening Past Concepts without Past Data: Class-Incremental Learning from Online Placebos
Yaoyao Liu, Yingying Li, Bernt Schiele, Qianru Sun
Knowledge Distillation Class Incremental Learning Unseen Entity Historical Data Class Data Placebo Effect

October 17, 2023

Rethinking Class-incremental Learning in the Era of Large Pre-trained Models via Test-Time Adaptation
Imad Eddine Marouf, Subhankar Roy, Enzo Tartaglione, Stéphane Lathuilière
Class Incremental Learning Test Time Adaptation New Era Large Pre Trained Model Transferable Representation

October 13, 2023

Federated Class-Incremental Learning with Prompting
Jiale Liu, Yu-Wei Zhan, Chong-Yu Zhang, Xin Luo, Zhen-Duo Chen, Yinwei Wei, Xin-Shun Xu
Catastrophic Forgetting Class Incremental Learning Federated Prompt Cooperation Tiny ImageNet Incremental Task

October 5, 2023

Class-Incremental Learning Using Generative Experience Replay Based on Time-aware Regularization
Zizhao Hu, Mohammad Rostami
Continual LEArning Continual Learning Class Incremental Learning Generative Replay Temporal Regularization

Class Incremental Learning

Papers

Learning Prompt with Distribution-Based Feature Replay for Few-Shot Class-Incremental Learning

Federated Class-Incremental Learning with New-Class Augmented Self-Distillation

Fine-Grained Knowledge Selection and Restoration for Non-Exemplar Class Incremental Learning

DSS: A Diverse Sample Selection Method to Preserve Knowledge in Class-Incremental Learning

PBES: PCA Based Exemplar Sampling Algorithm for Continual Learning

Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration

TLCE: Transfer-Learning Based Classifier Ensembles for Few-Shot Class-Incremental Learning

Class Incremental Learning for Adversarial Robustness

MIND: Multi-Task Incremental Network Distillation

Prompt-Based Exemplar Super-Compression and Regeneration for Class-Incremental Learning

Density Distribution-based Learning Framework for Addressing Online Continual Learning Challenges

From Categories to Classifiers: Name-Only Continual Learning by Exploring the Web

Combining Past, Present and Future: A Self-Supervised Approach for Class Incremental Learning

Class Incremental Learning with Pre-trained Vision-Language Models

Constructing Sample-to-Class Graph for Few-Shot Class-Incremental Learning

Look At Me, No Replay! SurpriseNet: Anomaly Detection Inspired Class Incremental Learning

Wakening Past Concepts without Past Data: Class-Incremental Learning from Online Placebos

Rethinking Class-incremental Learning in the Era of Large Pre-trained Models via Test-Time Adaptation

Federated Class-Incremental Learning with Prompting

Class-Incremental Learning Using Generative Experience Replay Based on Time-aware Regularization