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

A Hybrid Approach for Summarizing Text and Image Data Using ResNet and BART

  • Premanand Ghadekar,
  • Bijin Jiby,
  • Arunav Chandra,
  • Anveshika Kamble,
  • Rohit Arole,
  • Sukhpreet Bhatti,
  • Aditya Pratap Singh Kirar

摘要

With the increase in the volume and diversity of data available in various domains, the need for summarization techniques that can handle multiple data types has become crucial. Summarizing Heterogeneous data involves generating a concise representation of diverse data types, including text, image, audio, video, and structured data. In this paper, the focus is more on the major data types which are text, image, video, and text extracted from relational data. A multimodal architecture is created for processing and providing a summary in the text form. This involves a combined structure with several pre-trained models as well as Natural Language Processing-based libraries. Additionally, pre-trained models like BART and ResNet101 have been integrated.