ImplicitDeepfake: Plausible Face-Swapping through Implicit Deepfake Generation using NeRF and Gaussian Splatting ImplicitDeepfake:通过使用NeRF和高斯溅射的隐式Deepfake生成的合理换脸 Georgii Stanishevskii1∗ Jakub Steczk
Deepfake idea Motivation其他Semantic informationDataLoss functionsModel Motivation 考虑语义信息,在弱监督信息中尽可能挖掘有效信息;一种更贴近实际的实验设定;在更为精确的数据集上时(标注成本较高,比如为CUB200种鸟类标注,需要很高的专业知识水平,此时应用CLL),反标记监督信息更为有用: e.g.
Awesome-DeepFake-Learning The approach I work on DeepFake. Awesome-DeepFake-Learning 1. Intuitive Learning2. Survey Paper3. Curated lists4. Deepfakes Datasets5. Generation of synthetic content 5.1 G
文章目录 学习的博文资料1、DeepFake 入门了解2、深度“打假”:通过眨眼检测“deepfake”视频 换脸(Face-swapping)Original DeepfakesFaceswap-GAN3D face FaceSwapFast Face-swap Using CNN(ICCV2017)On Face Segmentation, Face Swapping, and Face