Summary of Fake News Detection and Manipulation Reasoning Via Large Vision-language Models, by Ruihan Jin et al.
Fake News Detection and Manipulation Reasoning via Large Vision-Language Models
by Ruihan Jin, Ruibo Fu, Zhengqi Wen, Shuai Zhang, Yukun Liu, Jianhua Tao
First submitted to arxiv on: 2 Jul 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a new multi-media research topic, manipulation reasoning, which aims to reason manipulations based on news content. To support this research, the authors introduce a benchmark for fake news detection and manipulation reasoning called Human-centric and Fact-related Fake News (HFFN). The HFFN benchmark highlights the centrality of human and high factual relevance, with detailed manual annotations. The paper also presents a Multi-modal news Detection and Reasoning langUage Model (M-DRUM) that not only judges the authenticity of multi-modal news but also raises analytical reasoning about potential manipulations. The M-DRUM model uses a cross-attention mechanism to extract fine-grained fusion features from multi-modal inputs and a large vision-language model (LVLM) as its backbone to facilitate fact-related reasoning. A two-stage training framework is deployed to better activate the capacity of identification and reasoning. Comprehensive experiments demonstrate that the proposed model outperforms state-of-the-art fake news detection models and powerful LVLMs like GPT-4 and LLaVA. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper tries to find ways to detect fake news and understand why people make it up. They want to develop a new way of analyzing news articles to figure out if they’re true or not, and also try to understand what’s behind the manipulation. To do this, they created a special dataset with lots of examples of real and fake news stories, and designed a special computer model that can analyze multiple types of data like text, images, and audio. The goal is to create a system that can accurately identify fake news and help us make sense of why people spread it. |
Keywords
» Artificial intelligence » Cross attention » Gpt » Language model » Multi modal
