Face swapping: automatically replacing faces in photographs

By D. Bitouk, Neeraj Kumar, S. Dhillon, Peter N. Belhumeur, Shree K. Nayar. ACM Transactions on Graphics, 2008

431 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/86dd113fe0b396384071504c6d562c90be3198b5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Fast Face-Swap Using Convolutional Neural Networks (2016, background): “[2], for example, automatically substituted an input face by another face selected from a large database of images based on the similarity of appearance and pose.”
Expression flow for 3D-aware face component transfer (2011, methodology): “The face swapping system [Bitouk et al. 2008] achieves a similar goal by constructing and using a large face image library.”
3D Morphable Face Models—Past, Present, and Future (2019, unknown): “source video. To this end, both persons are reconstructed based on the same parametric model resulting in dense inter-person correspondences. First approaches enabled face replacement between images [Bitouk et al. 2008; Blanz et al. 2004b; Jones et al. 2008; Kemelmacher-Shlizerman 2016]. Later works”
Learning to predict indoor illumination from a single image (2017, methodology): “Similar ideas have also been used in other 2D compositing techniques [Bitouk et al. 2008; Lalonde et al. 2007].”
On Face Segmentation, Face Swapping, and Face Perception (2017, methodology): “Some methods estimate 3D facial shapes [1, 4, 27, 28], by ﬁtting 3D Morphable Face Models (3DMM).”
DeepFakes: a New Threat to Face Recognition? Assessment and Detection (2018, background): “One of the ﬁrst works on face swapping is by Bitouk et al. [14], where the authors searched in a database for a face similar in appearance to the input face and then focused on perfecting the blending of the found face into the input image.”
I Know That Person: Generative Full Body and Face De-identification of People in Images (2017, unknown): “[4].”
Transfusive image manipulation (2012, methodology): “, using face detectors or automatic labeling based on learned features, photo manipulations such as tonal adjustments or object removal have been successfully demonstrated [Bitouk et al. 2008; Brand and Pletscher 2008; Berthouzoz et al. 2011], but a general pixel-accurate mapping between images as i”
Mask-Guided Portrait Editing With Conditional GANs (2019, methodology): “With the proposed structure that disentangles and recombines facial instance embeddings with face masks, our method also enhances over face swapping methods [5, 26] by supporting explicit face and hair swapping.”
Live Face De-Identification in Video (2019, unknown): “Face swapping, i.e., the replacement of a person’s face in an image with another person’s face, has been an active research topic for some time, starting with the inﬂuential work of [3, 2].”
Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics (2019, background): “Although in recent years there have been many sophisticated algorithms for generating realistic synthetic face videos [9, 13, 46, 51, 26, 47, 37, 20, 23, 10, 21, 50], most of these have not been in mainstream as open-source software tools that anyone can use.”
End-to-End Reconstruction-Classification Learning for Face Forgery Detection (2022, background): “The recent years have witnessed the considerable progress of face forgery generation methods [2,4,11,18,20, 41, 42, 50, 53].”
SimSwap: An Efficient Framework For High Fidelity Face Swapping (2020, methodology): “An automatic method [3] was proposed but could only swap faces with identities in a specific face library.”
Efficient privacy preserving video surveillance (2009, background): “Face swapping [5] and face deidentification [10, 13] try to modify face images such that they can be automatically detected, but cannot be correctly recognized.”
Search-and-replace editing for personal photo collections (2010, background): “Recent methods have exploited large unstructured internet-based image collections to search for the most compatible image region to blend in [18, 6].”