Scholarlink

<p>Comics are a one-of-a-kind style of visual storytelling that expertly weaves the complex web of human feelings via the depiction of the characters in the stories they tell. This research sets out on an intriguing quest to analyze and classify the emotions displayed by comic book characters. By utilizing cutting-edge deep learning (DL) technology, the proposed study can incorporate the powerful Visual Geometry Group version 16 (VGG16) model as a feature extractor, complicated attention mechanisms for weight optimization, and the EmotioNet model for multi-classification of emotional states. The results of this research shed light on the complex relationship between comics, the use of visual storytelling, and the expression of emotions. Notably, the proposed model can achieve a truly remarkable overall accuracy rate of 92.6% consistently. This demonstrates its expertise in identifying emotions across various comic tales and visual styles. The study makes a substantial contribution to the fields of computer vision (CV) and emotion recognition (ER). It also has deep implications for human–computer interaction and the analysis of multimedia content. It is a monument to the dynamic nature of technology, where creativity and precision work in harmony to decode the delicate language of human emotions within the alluring world of comics, and this achievement serves as a testament to the nature of technology.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deciphering emotions in comics: analysis of emotion classification of comic characters via attention-based deep learning models

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Rishabh Sharma,
  • Vinay Kukreja

摘要

Comics are a one-of-a-kind style of visual storytelling that expertly weaves the complex web of human feelings via the depiction of the characters in the stories they tell. This research sets out on an intriguing quest to analyze and classify the emotions displayed by comic book characters. By utilizing cutting-edge deep learning (DL) technology, the proposed study can incorporate the powerful Visual Geometry Group version 16 (VGG16) model as a feature extractor, complicated attention mechanisms for weight optimization, and the EmotioNet model for multi-classification of emotional states. The results of this research shed light on the complex relationship between comics, the use of visual storytelling, and the expression of emotions. Notably, the proposed model can achieve a truly remarkable overall accuracy rate of 92.6% consistently. This demonstrates its expertise in identifying emotions across various comic tales and visual styles. The study makes a substantial contribution to the fields of computer vision (CV) and emotion recognition (ER). It also has deep implications for human–computer interaction and the analysis of multimedia content. It is a monument to the dynamic nature of technology, where creativity and precision work in harmony to decode the delicate language of human emotions within the alluring world of comics, and this achievement serves as a testament to the nature of technology.