A Semantic Weight Adaptive Model Based on Visual Question Answering in Different Languages

Visual Question Answering (VQA) is a rapidly growing research field that combines computer vision and natural language processing to enable machines to understand visual content and answer questions related to images. This paper presents a multimodal web-based framework that integrates deep learning techniques for image understanding and language processing. The proposed system utilizes the BLIP-VQA model, which combines a Vision Transformer for extracting visual features and a transformer-based language model for generating contextual answers. The system is implemented using the Flask web framework and supports multilingual interaction through an integrated translation module that converts user queries into English and translates responses back into the user’s native language. The framework enables users to upload images and ask questions related to the visual content, receiving accurate responses in real time. Experimental results demonstrate that the proposed system effectively integrates multimodal learning and multilingual interaction, making it suitable for practical real-world applications.