MoviePoster-Grounded Contextual Visualization Using Multimodal Techniques

This study introduces an innovative approach to contextual visualization in movies called MoviePoster-Grounded Contextual Visualization (MPCV). Leveraging multimodal techniques, MPCV integrates textual and visual information, primarily focusing on movie posters, to enhance the exploration of movie-related datasets. By combining these modalities, MPCV seeks to reveal latent patterns and connections within movie datasets, offering a more comprehensive and contextually enriched visualization experience. To validate the effectiveness of MPCV, we conduct a thorough review of related work in movie poster analysis, genre classification, and recommendation systems. In particular, we have salvaged the advantages of pre-trained models such as MobileNetV3, Inception-v3, EfficientNetV2, and BLIP-2 to present a descriptive movie poster visualization.