Toru Ogawa

Sketch-based manga retrieval using manga109 dataset

Manga (Japanese comics) are popular worldwide. However, current e-manga archives offer very limited search support, i.e., keyword-based search by title or author. To make the manga search experience more intuitive, efficient, and enjoyable, we propose a manga-specific image retrieval system. The proposed system consists of efficient margin labeling, edge orientation histogram feature description with screen tone removal, and approximate nearest-neighbor search using product quantization. For querying, the system provides a sketch-based interface. Based on the interface, two interactive reranking schemes are presented: relevance feedback and query retouch. For evaluation, we built a novel dataset of manga images, Manga109, which consists of 109 comic books of 21,142 pages drawn by professional manga artists. To the best of our knowledge, Manga109 is currently the biggest dataset of manga images available for research. Experimental results showed that the proposed framework is efficient and scalable (70 ms from 21,142 pages using a single computer with 204 MB RAM).

Ion kinetic effect on bifurcated relaxation to a field-reversed configuration in TS-4 CT experiment

The ion kinetic effect on the bifurcated relaxation of merging spheromaks to a field-reversed configuration (FRC) was studied experimentally using varied S*, which is the ratio of the minor radius to the ion skin depth from 1 to 7. The two merging spheromaks were observed to relax to an FRC or a new spheromak depending on whether the initial poloidal eigenvalue was smaller or larger than a threshold value. The threshold initial poloidal eigenvalue for the relaxation to an FRC increased with the decreasing S* value. A decrease in S* promoted the relaxation to an FRC, annihilating the magnetic helicity, in sharp contrast with the conventional Taylor relaxation. Suppression of the low-n mode by the rotation shear of the toroidal modes is the most probable reason why the low-S* condition promotes the relaxation into an FRC.

ChainerCV: a Library for Deep Learning in Computer Vision

Despite significant progress of deep learning in the field of computer vision, there has not been a software library that covers these methods in a unifying manner. We introduce ChainerCV, a software library that is intended to fill this gap. ChainerCV supports numerous neural network models as well as software components needed to conduct research in computer vision. These implementations emphasize simplicity, flexibility and good software engineering practices. The library is designed to perform on par with the results reported in published papers and its tools can be used as a baseline for future research in computer vision. Our implementation includes sophisticated models like Faster R-CNN and SSD, and covers tasks such as object detection and semantic segmentation.

Manga109 dataset and creation of metadata

We have created Manga109, a dataset of a variety of 109 Japanese comic books publicly available for use for academic purposes. This dataset provides numerous comic images but lacks the annotations of elements in the comics that are necessary for use by machine learning algorithms or evaluation of methods. In this paper, we present our ongoing project to build metadata for Manga109. We first define the metadata in terms of frames, texts and characters. We then present our web-based software for efficiently creating the ground truth for these images. In addition, we provide a guideline for the annotation with the intent of improving the quality of the metadata.

Sketch simplification by classifying strokes

In this paper, we propose a novel approach to creating clean line drawing from a scribbled sketch automatically. The main problem is determining which strokes of a scribbled sketch should be merged. We use a machine learning approach to solve this problem. Our method can automatically generate training data by comparing scribbled sketches with manually drawn line drawings without using annotations. In order to verify the generated training data, we merged strokes and created clean line drawings in accordance with the generated training data. In addition, we trained a support vector machine to estimate the pairs of strokes to be merged. Further, we verified that our method can create line drawings using this estimator.