基于BLIP的VQA智能问答系统设计与实现
Design and Implementation of VQA Intelligent Q&A System Based on BLIP
艾孜买提江·买买提明 1杜亚洲 1麦合木提江·安外尔1
作者信息
- 1. 新疆第二医学院,新疆克拉玛依 834000
- 折叠
摘要
本论文介绍了基于BLIP(Bootstrapping Language Image Pre-training)模型的VQA(Visual Question Answering)智能问答系统的设计与实现.该系统通过将视觉和语言信息融合,实现了对图像提出问题并获取准确答案的功能.前端使用HTML、CSS和Javascript构建,后端采用Django框架进行开发,部署使用Docker容器.实验结果表明,所提出的系统在开放问答领域上取得了显著的成果.
Abstract
This thesis describes the design and implementation of VQA(Visual Question Answering)intelligent question and answer system based on BLIP(Bootstrapping Language Image Pre-training)model.The system realizes the function of asking questions about images and obtaining accurate answers by fusing visual and linguistic information.The front-end is built using HTML,CSS and Javascript,the back-end is developed using Django framework and deployed using Docker containers.The experimental results show that the proposed system achieves significant results on the open question and answer domain.
关键词
BLIP/VQA/智能问答系统/Django/DockerKey words
BLIP/VQA/intelligent question and answer system/django/docker引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据