Visual Narratives Unveiled: Comic Genre Classification through CNN-SVM Fusion

This study addresses the challenging issue of genre recognition in comics comprehensively by including CNN and SVM in a holistic manner. Identifying the genre of comics may be a challenging matter. The research includes an analysis of the model's recall, accuracy, and F1 score across many genres. You may assess the model's aptitude in identifying different forms of humorous material, and the table presents a concise overview of the study results. Through the use of this approach, the model showcases a high level of expertise in the specific field it is focusing on. Each genre, including historical dramas, sports, fantasy, romantic comedies, and wars, has its own set of distinct measurements. The measurements have three components: accuracy, recall, and F1 scores. The comic genre is consistently acknowledged, leading to the achievement of a notable algorithmic discriminative accuracy of 75.63 percent. The Love Romance model's recall rate of 79.20% indicates its ability to accurately recognize a substantial portion of real-world data related to the genre. The algorithm regularly produces favorable outcomes by using the F1 score, a complete metric that combines accuracy and recall. Certain forms of writing are more effective in illustrating these statistics than others. Comedy has an accuracy rate of 69.58%, whilst love romance has a precision rate of 76.25%. The global mean F1 score of 65.49 percent is a stable figure that stays unaffected by variations in genre.