Exploiting Style Latent Flows for Generalizing Deepfake Video Detection
Computing Research Repository (CoRR)(2024)
Chung-Ang University | NAVER | UNIST
Abstract
This paper presents a new approach for the detection of fake videos, based onthe analysis of style latent vectors and their abnormal behavior in temporalchanges in the generated videos. We discovered that the generated facial videossuffer from the temporal distinctiveness in the temporal changes of stylelatent vectors, which are inevitable during the generation of temporally stablevideos with various facial expressions and geometric transformations. Ourframework utilizes the StyleGRU module, trained by contrastive learning, torepresent the dynamic properties of style latent vectors. Additionally, weintroduce a style attention module that integrates StyleGRU-generated featureswith content-based features, enabling the detection of visual and temporalartifacts. We demonstrate our approach across various benchmark scenarios indeepfake detection, showing its superiority in cross-dataset andcross-manipulation scenarios. Through further analysis, we also validate theimportance of using temporal changes of style latent vectors to improve thegenerality of deepfake video detection.
MoreTranslated text
Key words
Deepfake Detection,Face Forgery Detection
PDF
View via Publisher
AI Read Science
AI Summary
AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.
Example
Background
Key content
Introduction
Methods
Results
Related work
Fund
Key content
- Pretraining has recently greatly promoted the development of natural language processing (NLP)
- We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
- We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
- The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
- Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance
Try using models to generate summary,it takes about 60s
Must-Reading Tree
Example

Generate MRT to find the research sequence of this paper
Related Papers
Mining Generalized Multi-timescale Inconsistency for Detecting Deepfake Videos
International Journal of Computer Vision 2024
被引用0
Data Disclaimer
The page data are from open Internet sources, cooperative publishers and automatic analysis results through AI technology. We do not make any commitments and guarantees for the validity, accuracy, correctness, reliability, completeness and timeliness of the page data. If you have any questions, please contact us by email: report@aminer.cn
Chat Paper
去 AI 文献库 对话