Class Impression for Data-free Incremental Learning

TL;DR

This paper introduces a data-free incremental learning method that synthesizes data from previous models to prevent forgetting, improving accuracy in classifying medical imaging data without storing real past data.

Contribution

It proposes a novel data-free framework using synthesized data and specialized loss functions to enhance incremental learning performance.

Findings

Improved accuracy on echocardiography classification tasks.

Effective mitigation of catastrophic forgetting without storing real data.

Outperforms state-of-the-art methods in class incremental learning.

Abstract

Standard deep learning-based classification approaches require collecting all samples from all classes in advance and are trained offline. This paradigm may not be practical in real-world clinical applications, where new classes are incrementally introduced through the addition of new data. Class incremental learning is a strategy allowing learning from such data. However, a major challenge is catastrophic forgetting, i.e., performance degradation on previous classes when adapting a trained model to new data. Prior methodologies to alleviate this challenge save a portion of training data require perpetual storage of such data that may introduce privacy issues. Here, we propose a novel data-free class incremental learning framework that first synthesizes data from the model trained on previous classes to generate a \ours. Subsequently, it updates the model by combining the synthesized data with new class data. Furthermore, we incorporate a cosine normalized Cross-entropy loss to mitigate the adverse effects of the imbalance, a margin loss to increase separation among previous classes and new ones, and an intra-domain contrastive loss to generalize the model trained on the synthesized data to real data. We compare our proposed framework with state-of-the-art methods in class incremental learning, where we demonstrate improvement in accuracy for the classification of 11,062 echocardiography cine series of patients.