Comic Book Images Research Articles

A learning approach with incomplete pixel-level labels for deep neural networks

Learning with incomplete labels in Neural Networks has been actively investigated these last years. Among different kinds of incomplete labels, we investigate incomplete pixel-level labels which are tackled in many concrete problems. One of the challenges for incomplete pixel-level labels is the missing information at local-level. Most of the current researches with incomplete labels in Neural Network focus on the incompleteness of global labels, only a few works focus on the incompleteness of local labels. To deal with the local incompleteness, we propose a learning approach which uses two dynamic weighted maps in parallel: one for object pixels and another one for background pixels. The two maps are integrated into the loss function of the target Neural Networks, to optimize the model by the present labels and to minimize the damage of the missing labels. We validate our approach on the speech balloon extraction problem in comic book images. Our approach uses the output of a balloon extraction algorithm as incomplete labels. The results are comparable with the state of the art supervised approach with manual labels. The results are very promising because our method does not require any manual labels. In addition, we apply our method to the medical image segmentation task to confirm the generalization of our approach.

Finding (One)self in Art Education: A Queer Conversation

Abstract This article, a two-person narrative, makes visible some challenges embedded in queering pre-service art education in an educational sphere that tends to be seen as heteronormative, White, able-bodied, and cisgendered—even for those who are queer themselves. In this piece, Ria, an undergraduate student, offers the reader a peek into a lived experience rarely shared as they describe the challenge of finding themselves at university, navigating the repeated coming out process both in and outside the classroom, undergoing treatment for cancer, and returning a changed person. The co-narrative offers the perspective of Ria’s university faculty member and mentor. Acknowledging moments of awareness and dissonance, she discusses the journey to rectify curriculum and pedagogy in a responsive manner while noting the value of being present for one’s student beyond the walls of the classroom. Ria’s comic book images are the vehicle through which life, grief, and moving forward are explored.

Comic MTL: optimized multi-task learning for comic book image analysis

Comic book image analysis methods often propose multiple algorithms or models for multiple tasks like panel and character (body and face) detection, balloon segmentation, text recognition, etc. In this work, we aim to reduce the processing time for comic book image analysis by proposing one model that can learn multiple tasks called Comic MTL instead of using one model per task. In addition to detection and segmentation tasks, we integrate the relation analysis task for balloons and characters into the Comic MTL model. The experiments are carried out on DCM772 and eBDtheque public datasets that contain the annotations for panels, balloons, characters and also the associations between balloon and character. We show that the Comic MTL model can detect the associations between balloons and their speakers (comic characters) and handle other tasks like panel and character detection and also balloons segmentation with promising results.

Digital Comics Image Indexing Based on Deep Learning

The digital comic book market is growing every year now, mixing digitized and digital-born comics. Digitized comics suffer from a limited automatic content understanding which restricts online content search and reading applications. This study shows how to combine state-of-the-art image analysis methods to encode and index images into an XML-like text file. Content description file can then be used to automatically split comic book images into sub-images corresponding to panels easily indexable with relevant information about their respective content. This allows advanced search in keywords said by specific comic characters, action and scene retrieval using natural language processing. We get down to panel, balloon, text, comic character and face detection using traditional approaches and breakthrough deep learning models, and also text recognition using LSTM model. Evaluations on a dataset composed of online library content are presented, and a new public dataset is also proposed.

Subgraph spotting in graph representations of comic book images

Graph-based representations are the most powerful data structures for extracting, representing and preserving the structural information of underlying data. Subgraph spotting is an interesting research problem, especially for studying and investigating the structural information based content-based image retrieval (CBIR) and query by example (QBE) in image databases. In this paper we address the problem of lack of freely available ground-truthed datasets for subgraph spotting and present a new dataset for subgraph spotting in graph representations of comic book images (SSGCI) with its ground-truth and evaluation protocol. Experimental results of two state-of-the-art methods of subgraph spotting are presented on the new SSGCI dataset.

‘A cross burning darkly, blackening the night’: Reading racialized spectacles of conflict and bondage in Marvel’s early Black Panther comics

Abstract All superheroes embody cultural messages and negotiate cultural conflicts. Yet black bodies, introduced to the New World as property and rigorously and often violently regulated by interconnected cultural and institutional forces ever since, have always occupied an especially vexed place in American culture. While all superheroes face repeated threats to their physical integrity, America’s first black superhero, Marvel’s Black Panther, created in 1966 by Stan Lee and Jack Kirby, faces the specific – and specifically risky – challenge of performing black embodiment without succumbing to the stereotypes of black embodiment, most notably the stereotype of black bodies being more bodily (re: more violent, more sexual, more animalistic) than white bodies. This article argues that although Black Panther’s early appearances in The Fantastic Four (1961–present), The Avengers (1963–present) and his first solo series, Jungle Action Featuring: The Black Panther (McGregor et al. 1973–76) exploit the superhero genre and comic book medium’s propensity for fantasy and iconic imagery to envision a blackness that is beautiful and liberating, depictions of Black Panther’s conflicts with racialized supervillains and animals as well as his routine depiction within gratuitous spectacles of suffering and bondage demonstrate a simultaneous – and occasionally overwhelming – tendency to appropriate the black body in the service of white desires and anxieties. Ultimately, this article asserts that Black Panther’s early stories prove both the power of comic book images and the danger of that power for subjects who are still fighting an uphill to be seen without being reduced to their visibility.

The Legend of Lightning and Thunder P. I. Rumbolt

Rumbolt, Paula Ikuutaq. The Legend of Lightning and Thunder. Iqualuit, NV: Inhabit Media, 2013. Print.There are many Inuit stories that tell of the origins of things. There are also many stories about orphans. This book is a simple retelling of a legend that involves both and carries a gentle lesson. It concludes that the presence of thunder and lightning in our world is the result of people neglecting and not caring for orphan children. Paula Ikuutaq Rumbolt is an Inuit writer from Baker Lake, Nunavut. She learned traditional Inuit stories and beliefs from her grandmother. The story is from “a time before stealing existed. No one knew what it was, as it had never happened to anyone.” When the orphan children are turned away from the camp, they must steal food to stay alive, which makes them permanent outcasts. They steal caribou meat, a small caribou skin and a flint. They discover that the skin will make noise and the flint will make sparks and have fun playing with them. They soon realize, though, that they will be hunted by the people in the camp so they hide in the sky, where you can hear them today, playing with the skin and the flint, making thunder and lightning. While the text is fairly easy to read, the presence of more difficult words such as “exhilarating” indicates that it is really intended to be shared by an adult and child.While the story is simple, it is brought to life by Jo Rioux’s artwork. For each pair of pages, one is a full page comic-book style image showing Inuit people in village settings and doing traditional activities. The facing page has a smaller image, with text usually printed over a light coloured sky. The images are in a palette of soft browns and oranges, which shade to reds, particularly where magical things are being depicted. Young children will enjoy this book, as will anyone who appreciates Inuit legends. This volume extends Inhabit Media's track record of consistently high quality publications and authentic voices and sources. Highly recommended for elementary school and public libraries.Highly recommended: 4 stars out of 4 Reviewer: Sandy CampbellSandy is a Health Sciences Librarian at the University of Alberta, who has written hundreds of book reviews across many disciplines. Sandy thinks that sharing books with children is one of the greatest gifts anyone can give.