A CNN-based Feature Space for Semi-supervised Incremental Learning in Assisted Living Applications
This addresses incremental learning challenges for CNNs in assisted living applications, but it is incremental as it builds on existing semi-supervised and feature space techniques.
The paper tackles the problem of CNNs struggling with new object instances in assisted living by proposing a semi-supervised incremental learning method that uses feature space information to label problematic images, resulting in a 40% improvement in classification accuracy for new instances.
A Convolutional Neural Network (CNN) is sometimes confronted with objects of changing appearance ( new instances) that exceed its generalization capability. This requires the CNN to incorporate new knowledge, i.e., to learn incrementally. In this paper, we are concerned with this problem in the context of assisted living. We propose using the feature space that results from the training dataset to automatically label problematic images that could not be properly recognized by the CNN. The idea is to exploit the extra information in the feature space for a semi-supervised labeling and to employ problematic images to improve the CNN's classification model. Among other benefits, the resulting semi-supervised incremental learning process allows improving the classification accuracy of new instances by 40% as illustrated by extensive experiments.