arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2310.11204cs.CVeess.IV
收藏

改进视频深度伪造检测:一种基于DCT的块级分析方法

Improving Video Deepfake Detection: A DCT-Based Approach with Patch-Level Analysis

University of Catania(卡塔尼亚大学)

Luca Guarnera, Salvatore Manganello, Sebastiano Battiato

AI总结:

提出一种基于DCT和I帧块级分析的视频深度伪造检测算法,发现眼睛和嘴巴区域最具判别力,显著提升了检测效率与准确性。

摘要:

本文提出了一种用于检测数字视频中深度伪造的新算法。为了提供比文献中描述的方法更快的计算和分析速度,提取了I帧。为了识别单个视频帧内的判别区域,分别对整个帧、背景、人脸、眼睛、鼻子、嘴巴和人脸边框进行了分析。从离散余弦变换(DCT)中,从AC系数中提取了Beta分量,并用作标准分类器的输入。实验结果表明,眼睛和嘴巴区域最具判别力,能够确定所分析视频的真伪性质。

英文摘要:

A new algorithm for the detection of deepfakes in digital videos is presented. The I-frames were extracted in order to provide faster computation and analysis than approaches described in the literature. To identify the discriminating regions within individual video frames, the entire frame, background, face, eyes, nose, mouth, and face frame were analyzed separately. From the Discrete Cosine Transform (DCT), the Beta components were extracted from the AC coefficients and used as input to standard classifiers. Experimental results show that the eye and mouth regions are those most discriminative and able to determine the nature of the video under analysis.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑