Transcending Forgery Specificity with Latent Space Augmentation for Generalizable Deepfake Detection
CVPR, 2024
Zhiyuan Yan（晏志远）
Postgraduate Student
The Chinese University of Hong Kong, Shenzhen (CUHKSZ)
Email / Github / Linkedin / Scholar / OpenReview
Biography
I am currently a postgraduate student at the Chinese University of Hong Kong, Shenzhen (CUHKSZ) under the esteemed guidance of Prof. Baoyuan Wu. I am also honored to be in active cooperation with Prof. Siwei Lyu and Dr. Yong Zhang.
Before this, I gained research and engineering experience at Tencent, SCLBD, Baidu, and DJI. My research primarily focuses on AIGC/Face Forgery detection, AI4Science, and AI Security.
News
- [April 2024] The paper on using ChatGPT-4V for detecting AI-synthesized images is accepted by CVPR 2024 Workshop on Media Forensics (WMF).
- [February 2024] The paper of LSDA has been accepted by CVPR 2024: Augmenting different forgery types in the latent space to address the forgery-specific overfitting problem in deepfake detection.
- [February 2024] I am invited as a reviewer of ECCV 2024.
- [October 2023] I am invited as a reviewer of CVPR 2024.
- [October 2023] I will go to Paris to attend ICCV 2023.
- [September 2023] Our DeepfakeBench has been accepted by NeurIPS 2023.
- [July 2023] The paper of UCF has been accepted by ICCV 2023: A new disentanglement-based framework for more general deepfake detection.
- [July 2023] Our DeepfakeBench, a comprehensive open-source platform for deepfake detection, is now available!
- [May 2022] Our HelixADMET has been published at Bioinformatics journal: a robust ADMET system for accurate drug screening and estimation of ADMET endpoints.
Experience
-
Tencent Youtu Lab
Jan 2024 - Now
Position: Research Intern for AIGC/Face Forgery Detection
-
Tencent AILab & SCLBD
Nov 2022 - Dec 2023
Position: Research Intern for AIGC/Face Forgery Detection
-
Baidu Co., Ltd
Sep 2021 - Jun 2022
Position: Research Intern for AI4Science
-
DJI Technology Co., Ltd.
Jan 2021 - Jun 2021
Position: Engineering Intern for Object Detection and Segmentation
Research Highlights (🧑💻 Co-first Author, 📮 Corresponding Author)
DeepfakeBench: A Comprehensive Benchmark of Deepfake Detection
NeurIPS, 2023
UCF: Uncovering Common Features for Generalizable Deepfake Detection
ICCV, 2023
HelixADMET: a robust and endpoint extensible ADMET system incorporating self-supervised knowledge transfer
Shanzhuo Zhang, Zhiyuan Yan, Yueyang Huang, lihang liu, Donglong He, Wei Wang, Xiaomin Fang 📮, Xiaonan Zhang, Fan Wang 📮, Hua Wu, and Haifeng Wang.
Bioinformatics, 2022
Competitions
- I have won the Competition Expert title in the Kaggle competition platform. (see my profile)
- Kaggle competition (NLP): Silver medal of Commonlit Readability prize (Global Top 5%)
- Kaggle competition (CV): Bronze medal of Hubmap Hacking the Kidney (Global Top 10%)
- Third Prize in the Nvidia 2021 TensorRT Accelerated AI Reasoning National Developer Competition (Top 5 in China)
- 2021 & 2020 Best Presentation in the China Undergraduate Physics Tournament (CUPT) (Top 1 in China)
Contact