Abstract
Liveliness detection is increasingly important in video conferencing due to the advent of deepfakes and their use in conducting online fraud. Currently, liveliness detection technologies lag behind deepfake creation technologies. This submission proposes the use of the bidirectional nature of video conferencing to counter deepfakes more reliably and more efficiently than the state-of-the-art.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Godavarti, Mahesh; Mouline, Ali; Wojcicki, Kamil; Achanta, Radhakrishna; and Casas, Raul, "REAL-TIME DEEPFAKE DETECTION DURING VIDEO CONFERENCES: HARNESSING BIDIRECTIONAL AUDIO-VIDEO FEATURES FROM REFLECTIONS AND ECHOES", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/8929