OpenIncrement: A Unified Framework for Open Set Recognition and Deep Class-Incremental Learning
This addresses the issue of degraded model performance in practical deep classifiers for incremental learning, though it appears incremental as it combines existing techniques.
The paper tackles the problem of misidentifying novel samples in deep incremental learning by integrating open set recognition, resulting in a framework that outperforms state-of-the-art incremental learning techniques and shows superior open set recognition performance compared to baselines.
In most works on deep incremental learning research, it is assumed that novel samples are pre-identified for neural network retraining. However, practical deep classifiers often misidentify these samples, leading to erroneous predictions. Such misclassifications can degrade model performance. Techniques like open set recognition offer a means to detect these novel samples, representing a significant area in the machine learning domain. In this paper, we introduce a deep class-incremental learning framework integrated with open set recognition. Our approach refines class-incrementally learned features to adapt them for distance-based open set recognition. Experimental results validate that our method outperforms state-of-the-art incremental learning techniques and exhibits superior performance in open set recognition compared to baseline methods.