Learning Self-Consistency for Deepfake Detection
ICCV 2021 · 538 citations
Detects deepfakes by learning to extract and compare source-feature inconsistencies within forged images, achieving state-of-the-art cross-dataset generalization.
Researcher spanning deepfake detection, face anti-spoofing, generative modeling, neural quantum states, and scientific machine learning. Work published at ICCV, ICLR, CVPR, WACV, SC, and in journals including Quantitative Finance, Machine Learning: Science and Technology, and Quantum Machine Intelligence.
Methods for detecting face forgeries and securing biometric authentication systems, from self-consistency learning and language-guided detection to robust anti-spoofing system design and source-free domain adaptation.
ICCV 2021 · 538 citations
Detects deepfakes by learning to extract and compare source-feature inconsistencies within forged images, achieving state-of-the-art cross-dataset generalization.
arXiv 2024 · 10 citations
Presents a comprehensive taxonomy of face anti-spoofing threats and proposes design principles for building robust systems resilient to digital injection, deepfake, and replay attacks.
CVPR 2025 · 8 citations
Enables privacy-preserving test-time adaptation of face anti-spoofing models via optimal transport-guided prototype matching and geodesic mixup augmentation.
WACV 2025 · 2 citations
Integrates language guidance and commonsense reasoning into deepfake detection via vision-language contrastive learning, achieving state-of-the-art generalization across unseen forgery methods.
Advances in conditional generation, adversarial robustness, generative data augmentation, model diagnosis, and efficient vision-language adaptation across diverse visual domains.
ICLR 2019 · 268 citations
Proposes a simple diversity-sensitive regularization for conditional GANs that explicitly encourages diverse outputs, solving mode collapse across image translation, inpainting, and video prediction.
ICCV 2019 · 108 citations
Trains a stochastic generator to produce diverse adversarial attacks, yielding more robust defenses than single-attack adversarial training.
arXiv 2025 · 1 citation
Proposes a salient concept-aware generative augmentation framework that disentangles class-discriminative features from irrelevant context, improving fine-grained recognition under both conventional and long-tail settings.
CVPR 2025 · 1 citation
Automates model diagnosis and correction by analyzing error patterns, testing causal hypotheses via attribute editing, and generating counterfactual training samples for improved robustness.
arXiv 2026 · 0 citations
Introduces CRAFT, a discrete-codebook method for adapting vision encoders in LVLMs to domain-specific tasks without modifying the language model, enabling portable encoder adaptation across architectures.
Scalable neural-network wavefunctions for quantum chemistry and combinatorial optimization, GPU-parallel variational Monte Carlo, natural evolution strategies for quantum systems, and meta-learning for transferable quantum state optimization.
Machine Learning: Science and Technology, 2023 · 65 citations
Introduces GPU-scalable parallelization strategies for neural-network variational Monte Carlo applied to ab-initio quantum chemistry, achieving systematic speedups over existing neural-network methods.
Machine Learning: Science and Technology, 2021 · 32 citations
Introduces quantum natural evolution strategies as a geometric unification of quantum/classical black-box optimization algorithms, achieving competitive Max-Cut approximation ratios.
SC 2021 · 27 citations
Replaces MCMC with autoregressive exact sampling in variational quantum Monte Carlo, enabling GPU-scalable parallelism for up to ten-thousand dimensional problems.
arXiv 2020 · 8 citations
Applies natural evolution strategies to quantum approximate optimization, achieving competitive Max-Cut approximation ratios via neural quantum states.
arXiv 2020 · 4 citations
Connects meta-learning to quantum ground-state problems, proposing Meta Variational Monte Carlo that accelerates training and improves convergence on random Max-Cut instances.
Quantum Machine Intelligence, 2023 · 0 citations
Journal version presenting Meta Variational Monte Carlo, which applies meta-learning to accelerate neural quantum state optimization across random Hamiltonian ensembles.
PhD Thesis, University of Michigan, 2022
PhD thesis synthesizing neural quantum states for scientific computing, covering variational Monte Carlo for quantum chemistry, combinatorial optimization, and financial derivative pricing.
Extending variational quantum Monte Carlo beyond physics to solve time-dependent partial differential equations, with applications to multi-asset derivative pricing in quantitative finance.
Quantitative Finance, 2024 · 16 citations
Generalizes variational quantum Monte Carlo with neural-network quantum states to solve arbitrary time-dependent PDEs, demonstrated on multi-asset Black-Scholes option pricing.