“IllustStyleMap: Visualization of Illustrations based on Similarity of Drawing Style of Authors” by Matsumoto, Matsui, Yamasaki, Aizawa and Katagiri
Next: “ILM Facial Performance Capture” »
« Previous: “Illustrative Visualization for Medicine and...
Conference:
Type:
Entry Number: 54
Title:
- IllustStyleMap: Visualization of Illustrations based on Similarity of Drawing Style of Authors
Presenter(s)/Author(s):
Abstract:
Illustration is one of the most familiar forms for people to express their mind and communicate with each other. A lot of illustrations drawn by novices or hobbyists are viewed and shared on illustration-sharing SNSs. (Pixiv1 is one of the most popular services.) However, how to retrieve and display illustrations has received relatively sparse attention. In conventional retrieval interface, we can retrieve illustrations only by keywords and tags, and results are always displayed in a grid layout (Figure 2) .
We propose IllustStyleMap, a method of retrieving and visualizing illustrations based on similarity of drawing style of authors, which is beneficial to illustration-sharing SNSs.
References:
- Douze, M., Ramisa, A., and Schmid, C. 2011. Combining attributes and fisher vectors for efficient image retrieval. In CVPR, 745–752.
- van der Maaten, L., and Hinton, G. 2008. Visualizing data using t-sne. JMLR 9, 2579–2605.
Additional Images:
