TT-DF: A Large-Scale Diffusion-Based Dataset and Benchmark for Human Body Forgery Detection

Wenkui Yang, Zhida Zhang, Xiaoqiang Zhou, Junxian Duan, Jie Cao ( 🖂 ) {}^{(\textrm{\Letter})} · 2025

Read TT-DF: A Large-Scale Diffusion-Based Dataset and Benchmark for Human Body Forgery Detection by Wenkui Yang, Zhida Zhang, Xiaoqiang Zhou, Junxian Duan, Jie Cao ( 🖂 ) {}^{(\textrm{\Letter})} (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-08437.txt.

Description

Wenkui Yang and co-authors build a large data set for detecting forged human body videos in a 2025 paper. Face deepfake work has many data sets and detectors, which eased security worries for faces. Body forgery lagged, with no large data set and few detectors, due to later starts and harder generation methods.

Their TikTok DeepFake set holds thousands of forged videos and well over a million synthetic frames, made with several advanced animation models. Two generation setups split identity and pose in different ways, and several compressed versions mimic data in the wild. A benchmark is included, plus a detector that uses motion flow patterns over time to spot fakes.

Catalog record

CREATORWenkui Yang, Zhida Zhang, Xiaoqiang Zhou, Junxian Duan, Jie Cao ( 🖂 ) {}^{(\textrm{\Letter})}
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-08437

Browse the full library · About the Tower of Babel