Clustering of illustrations by atmosphere using a combination of supervised and unsupervised learning
This work addresses the challenge of categorizing illustrations by atmosphere for improved recommendations and searches on platforms like Twitter and Pixiv, representing an incremental advancement in image clustering.
The paper tackled the problem of classifying illustrations by their elusive 'atmosphere' for social media applications by combining supervised and unsupervised learning with pseudo-labels, resulting in a method that outperformed conventional approaches in human-like clustering on manually classified datasets.
The distribution of illustrations on social media, such as Twitter and Pixiv has increased with the growing popularity of animation, games, and animated movies. The "atmosphere" of illustrations plays an important role in user preferences. Classifying illustrations by atmosphere can be helpful for recommendations and searches. However, assigning clear labels to the elusive "atmosphere" and conventional supervised classification is not always practical. Furthermore, even images with similar colors, edges, and low-level features may not have similar atmospheres, making classification based on low-level features challenging. In this paper, this problem is solved using both supervised and unsupervised learning with pseudo-labels. The feature vectors are obtained using the supervised method with pseudo-labels that contribute to an ambiguous atmosphere. Further, clustering is performed based on these feature vectors. Experimental analyses show that our method outperforms conventional methods in human-like clustering on datasets manually classified by humans.