Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics

TLDR

Deep‑fake face‑swap videos pose a growing threat to online trust, yet existing datasets are low‑quality and unrepresentative of real‑world DeepFakes. This work introduces Celeb‑DF, a large‑scale dataset of 5,639 high‑quality celebrity DeepFakes. Celeb‑DF was generated using an improved synthesis pipeline and is accompanied by a systematic evaluation of state‑of‑the‑art detection methods. The evaluation demonstrates that Celeb‑DF presents a markedly higher level of difficulty for current DeepFake detectors.

Abstract

AI-synthesized face-swapping videos, commonly known as DeepFakes, is an emerging problem threatening the trustworthiness of online information. The need to develop and evaluate DeepFake detection algorithms calls for datasets of DeepFake videos. However, current DeepFake datasets suffer from low visual quality and do not resemble DeepFake videos circulated on the Internet. We present a new large-scale challenging DeepFake video dataset, Celeb-DF, which contains 5,639 high-quality DeepFake videos of celebrities generated using improved synthesis process. We conduct a comprehensive evaluation of DeepFake detection methods and datasets to demonstrate the escalated level of challenges posed by Celeb-DF.