Leveraging Old Knowledge to Continually Learn New Classes in Medical Images

2023-03-24 · Evelyn Chee, Mong Li Lee, Wynne Hsu

Class-incremental continual learning is a core step towards developing artificial intelligence systems that can continuously adapt to changes in the environment by learning new concepts without forgetting those previously learned. This is especially needed in the medical domain where continually learning from new incoming data is required to classify an expanded set of diseases. In this work, we focus on how old knowledge can be leveraged to learn new classes without catastrophic forgetting. We propose a framework that comprises of two main components: (1) a dynamic architecture with expanding representations to preserve previously learned features and accommodate new features; and (2) a training procedure alternating between two objectives to balance the learning of new features while maintaining the model's performance on old classes. Experiment results on multiple medical datasets show that our solution is able to achieve superior performance over state-of-the-art baselines in terms of class accuracy and forgetting.

📄 PDF Abstract BibTeX arXiv:2303.13752

Code (1)

evelynchee/lo2ln 공식 구현 pytorch

Tasks

Continual Learning

FedProK: Trustworthy Federated Class-Incremental Learning via Prototypical Feature Knowledge Transfer

2024-05-04 · Xin Gao, Xin Yang, Hao Yu, Yan Kang 외

Federated Class-Incremental Learning (FCIL) focuses on continually transferring the previous knowledge to learn new classes in dynamic Federated Learning (FL). However, existing methods do not consider the trustworthines…

class-incremental learning Class Incremental Learning Federated Learning Incremental Learning+1

Few-Shot Incremental Learning with Continually Evolved Classifiers

2021-04-07 · CVPR 2021 1 · Chi Zhang, Nan Song, Guosheng Lin, Yun Zheng 외

Few-shot class-incremental learning (FSCIL) aims to design machine learning algorithms that can continually learn new concepts from a few data points, without forgetting knowledge of old classes. The difficulty lies in t…

class-incremental learning Class Incremental Learning Few-Shot Class-Incremental Learning Incremental Learning

Injecting Structured Biomedical Knowledge into Language Models: Continual Pretraining vs. GraphRAG

2026-04-03 · Jaafer Klila, Sondes Bannour Souihi, Rahma Boujelben, Nasredine Semmar 외 arxiv

The injection of domain-specific knowledge is crucial for adapting language models (LMs) to specialized fields such as biomedicine. While most current approaches rely on unstructured text corpora, this study explores two…

Continual Pretraining Question Answering

Constrained Few-shot Class-incremental Learning

2022-03-30 · CVPR 2022 1 · Michael Hersche, Geethan Karunaratne, Giovanni Cherubini, Luca Benini 외

Continually learning new classes from fresh data without forgetting previous knowledge of old classes is a very challenging research problem. Moreover, it is imperative that such learning must respect certain memory and …

class-incremental learning Class Incremental Learning continual few-shot learning Few-Shot Class-Incremental Learning+1

Continual Zero-Shot Learning through Semantically Guided Generative Random Walks

2023-08-23 · ICCV 2023 1 · Wenxuan Zhang, Paul Janson, Kai Yi, Ivan Skorokhodov 외

Learning novel concepts, remembering previous knowledge, and adapting it to future tasks occur simultaneously throughout a human's lifetime. To model such comprehensive abilities, continual zero-shot learning (CZSL) has …

Novel Concepts Zero-Shot Learning