CNN-Based Joint Clustering and Representation Learning with Feature Drift Compensation for Large-Scale Image Data
This work addresses the challenge of scalable image clustering for applications like data organization, but it is incremental as it builds on existing clustering and representation learning techniques.
The authors tackled the problem of efficiently clustering large unlabeled image sets by proposing a CNN-based method that jointly learns representations and performs clustering with feature drift compensation, achieving improved accuracy and storage complexity on datasets with millions of images.
Given a large unlabeled set of images, how to efficiently and effectively group them into clusters based on extracted visual representations remains a challenging problem. To address this problem, we propose a convolutional neural network (CNN) to jointly solve clustering and representation learning in an iterative manner. In the proposed method, given an input image set, we first randomly pick k samples and extract their features as initial cluster centroids using the proposed CNN with an initial model pre-trained from the ImageNet dataset. Mini-batch k-means is then performed to assign cluster labels to individual input samples for a mini-batch of images randomly sampled from the input image set until all images are processed. Subsequently, the proposed CNN simultaneously updates the parameters of the proposed CNN and the centroids of image clusters iteratively based on stochastic gradient descent. We also proposed a feature drift compensation scheme to mitigate the drift error caused by feature mismatch in representation learning. Experimental results demonstrate the proposed method outperforms start-of-the-art clustering schemes in terms of accuracy and storage complexity on large-scale image sets containing millions of images.