Competitive Learning Enriches Learning Representation and Accelerates the Fine-tuning of CNNs
This addresses the data labeling bottleneck for CNNs by enabling more efficient use of unlabeled data, though it appears incremental as it combines existing competitive learning with standard CNN architecture.
The authors tackled the problem of CNNs requiring large amounts of labeled data by integrating unsupervised competitive learning into convolutional layers to leverage unlabeled data for representation learning. Their validation experiments showed this approach effectively extracted image bases into filters and accelerated fine-tuning speed, with error rates decreasing steeply in initial phases when using many filters.
In this study, we propose the integration of competitive learning into convolutional neural networks (CNNs) to improve the representation learning and efficiency of fine-tuning. Conventional CNNs use back propagation learning, and it enables powerful representation learning by a discrimination task. However, it requires huge amount of labeled data, and acquisition of labeled data is much harder than that of unlabeled data. Thus, efficient use of unlabeled data is getting crucial for DNNs. To address the problem, we introduce unsupervised competitive learning into the convolutional layer, and utilize unlabeled data for effective representation learning. The results of validation experiments using a toy model demonstrated that strong representation learning effectively extracted bases of images into convolutional filters using unlabeled data, and accelerated the speed of the fine-tuning of subsequent supervised back propagation learning. The leverage was more apparent when the number of filters was sufficiently large, and, in such a case, the error rate steeply decreased in the initial phase of fine-tuning. Thus, the proposed method enlarged the number of filters in CNNs, and enabled a more detailed and generalized representation. It could provide a possibility of not only deep but broad neural networks.