Teacher-Student Structure for Domain Adaptation in Ensemble Audio-Visual Video Deepfake Detection
Elham Abolhasani, Maryam Ramezani, Hamid R. Rabiee
arXiv|2026|Preprint
Research
Trustworthy AI and Misinformation
Detecting manipulated content, labelling news early, and making predictions that report their own uncertainty.
Machine learning systems deployed on social platforms are adversarial by nature: the content they judge is produced by people who know they are being judged. We work on detection under that pressure — multi-modal fake news detection across text and images, audio-visual deepfake detection that generalises to unseen domains, and labelling news as early in its propagation as accuracy allows.
A second thread runs through all of it: a model should know what it does not know. We study epistemic uncertainty in prediction, so that a system can abstain instead of asserting.
Publications in this area
Epistemic Uncertainty-aware Recommendation Systems via Bayesian Deep Ensemble Learning
Radin Cheraghi, Amir Mohammad Mahfoozi, Sepehr Zolfaghari, Mohammadshayan Shabani, Maryam Ramezani, Hamid R. Rabiee
arXiv|2025|Preprint
FNR: A Similarity and Transformer-Based Approach to Detect Multi-Modal Fake News in Social Media
Faeze Ghorbanpour, Maryam Ramezani, Mohammad Amin Fazli, Hamid R. Rabiee
Social Network Analysis and Mining|2023
News Labeling as Early as Possible: Real or Fake?
Maryam Ramezani, Mina Rafiei, Soroush Omranpour, Hamid R. Rabiee
IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining|2019
