Object Detection for Comics using Manga109 Annotations

TL;DR

This paper introduces Manga109-annotations, a large-scale comic dataset, and SSD300-fork, a CNN-based detector tailored for comic object detection, addressing challenges of overlapping objects and lack of data.

Contribution

The paper provides the first large-scale annotated comic dataset and a novel CNN model optimized for comic object detection, improving accuracy over existing methods.

Findings

SSD300-fork outperforms other detection methods on Manga109-annotations

The new dataset enables better training for comic object detection

Overlapping objects in comics pose unique challenges addressed by the model

Abstract

With the growth of digitized comics, image understanding techniques are becoming important. In this paper, we focus on object detection, which is a fundamental task of image understanding. Although convolutional neural networks (CNN)-based methods archived good performance in object detection for naturalistic images, there are two problems in applying these methods to the comic object detection task. First, there is no large-scale annotated comics dataset. The CNN-based methods require large-scale annotations for training. Secondly, the objects in comics are highly overlapped compared to naturalistic images. This overlap causes the assignment problem in the existing CNN-based methods. To solve these problems, we proposed a new annotation dataset and a new CNN model. We annotated an existing image dataset of comics and created the largest annotation dataset, named Manga109-annotations. For the assignment problem, we proposed a new CNN-based detector, SSD300-fork. We compared SSD300-fork with other detection methods using Manga109-annotations and confirmed that our model outperformed them based on the mAP score.