This repository provides a list a list of publications from DASHLab about Deepfakes.
Our published datasets of deepfakes.
2021: FakeAVCeleb: A Novel Audio-Video Multimodal Deepfake Dataset accepted by NeurIPS21. [paper]
2022: Zoom-DF: A dataset for Video Conferencing Deepfake accepted by WDC22. [paper ]
Methods to generalize deepfake detectors.
2021: One Detector to Rule Them All: Towards a General Deepfake Attack Detection Framework accepted by WWW21. [paper]
2021: FReTAL: Generalizing Deepfake Detection using Knowledge Distillation and Representation Learning accepted by CVPRW21. [paper]
2021: TAR: Generalized Forensic Framework to Detect Deepfakes using Weakly Supervised Learning accepted by IFIP-SEC 21. [paper]
Our published papers that related to detecting low-quality compressed deepfakes.
2022: ADD: Frequency Attention and Multi-View Based Knowledge Distillation to Detect Low-Quality Compressed Deepfake Images accepted by AAAI22. [paper]
2022: BZNet: Unsupervised Multi-scale Branch Zooming Network for Detecting Low-quality Deepfake Videos accepted by WWW22. [paper]
Our published papers that related to applying continual learning to detect newly arrived deepfakes/GAN.
2020: T-GD: Transferable GAN-generated Images Detection Framework accepted by ICML20. [paper]
2021: CoReD: Generalizing Fake Media Detection with Continual Representation using Distillation accepted by ACMMM21. [paper]
Our published papers that related to evasion techniques to fool deepfake detectors.
2022: Am I a Real or Fake Celebrity? Evaluating Face Recognition and Verification APIs under Deepfake Impersonation Attack accepted by WWW2022. [paper]
2022: Evading Deepfake Detectors via High-Quality Face Pre-Processing Methods accepted by ICPR22. [paper]
2022: Negative Adversarial Example Generation Against Naver’s Celebrity Recognition API accepted by WDC22. [paper]
2022: A Face Pre-Processing Approach to Evade Deepfade Detector accepted by WDC22. [paper]
2023: Evaluating Racial Bias in Face Recognition APIs using Deepfakes accepted by IEEE Computer Magazine.
Our published papers that related to applying multi model to detect deepfakes.
2022: Evaluation of an Audio-Video Multimodal Deepfake Dataset using Unimodal and Multimodal Detectors accepted by ADGD21. [paper]
2019: GAN is a Friend or Foe? A Framework to Detect Various Fake Face Images accepted by ACMSAC19. [paper]
2020: OC-FakeDect: Classifying Deepfakes Using One-class Variational Autoencoder accepted by CVPRW20. [paper]
2020: FDFtNet: Facing Off Fake Images using Fake DetectionFine-tuning Network accepted by IFIP-SEC 20. [paper]
2021: Exploring the Asynchronous of the Frequency Spectra of GAN-generated Facial Images accepted by IJCAIW21. [paper]
2022: Deepfake Detection for Facial Images with Facemasks accepted by WDC22. [paper]