TT-BLIP: Enhancing Fake News Detection Using BLIP and Tri-Transformer

TL;DR

This paper presents TT-BLIP, an end-to-end multimodal model that combines vision and language pretraining with tri-transformer architecture to improve fake news detection accuracy.

Contribution

Introduces TT-BLIP, a novel integrated multimodal fake news detection model leveraging BLIP pretraining and tri-transformer fusion for enhanced multimodal understanding.

Findings

Outperforms state-of-the-art models on Weibo dataset

Achieves higher accuracy on Gossipcop dataset

Demonstrates effectiveness of multimodal fusion in fake news detection

Abstract

Detecting fake news has received a lot of attention. Many previous methods concatenate independently encoded unimodal data, ignoring the benefits of integrated multimodal information. Also, the absence of specialized feature extraction for text and images further limits these methods. This paper introduces an end-to-end model called TT-BLIP that applies the bootstrapping language-image pretraining for unified vision-language understanding and generation (BLIP) for three types of information: BERT and BLIPTxt for text, ResNet and BLIPImg for images, and bidirectional BLIP encoders for multimodal information. The Multimodal Tri-Transformer fuses tri-modal features using three types of multi-head attention mechanisms, ensuring integrated modalities for enhanced representations and improved multimodal data analysis. The experiments are performed using two fake news datasets, Weibo and Gossipcop. The results indicate TT-BLIP outperforms the state-of-the-art models.