About Me

  • I’m Yun-Yun (Alice) Tsai (蔡昀芸), a third year Ph.D. candidate in the Department of Computer Science at Columbia University, advised by Professor Junfeng Yang.
  • My research interests focus on Security in Artificial Intelligence, which I am particularly interested in improving trustworthy, security, and robustness over machine learning (ML) algorithms and computer systems.
  • I received M.S. and B.S. in computer science, both from National Tsing Hua University (NTHU), Taiwan. Previously, I was advised by Professor Tsung-Yi Ho and Dr. Pin-Yu Chen from IBM Research Trusted AI group.

News

  • [2024 May]: I will start my research scientist intern at Meta GenAI Team @ New York, NY.
  • [2024 April]: I passed my Ph.D. candidacy exam on April 24 2024. Thank to my committee members Prof. Richard Zemel, Prof. Carl Vondrick, and Prof. Junfeng Yang. (Slides)
  • [2024 Feb]: One main conf. paper + two workshop papers are accepted by CVPR 2024. See you in Seattle!
  • [2023 Sep]: One paper is accepted by NeurIPS 2023 @ New Orleans.
  • [2023 May]: Started my internship as an applied scientist at Amazon Astro Team at Bellevue Washington.
  • [2023 Mar]: Two papers are accepted by CVPR 2023 @ Vancouver.

Publications

§ Conference and Workshop Papers

  1. GDA: Generalized Diffusion for Robust Test-time Adaptation (Paper)
    • Yun-Yun Tsai, Fu-Chen Chen, Albert Y. C. Chen, Junfeng Yang, Che-Chun Su, Min Sun, Cheng-Hao Kuo
    • The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), 2024
  2. From Detection to Deception: Are AI-Generated Image Detectors Adversarially Robust? (Paper)
    • Yun-Yun Tsai, Ruize Xu, Chengzhi Mao, Junfeng Yang
    • The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), Responsible Generative AI Workshop, 2024
  3. Towards Robust Detection of AI-Generated Videos (Paper)
    • Qingyuan Liu, Pengyuan Shi, Yun-Yun Tsai, Chengzhi Mao, Junfeng Yang
    • The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), Generative Models for Computer Vision Workshop, 2024
  4. Convolutional Visual Prompt for Robust Visual Perception (Paper)
    • Yun-Yun Tsai, Chengzhi Mao, Junfeng Yang
    • 37th Conference on Neural Information Processing Systems (NeurIPS) 2023
  5. Towards Compositional Adversarial Robustness: Generalizing Adversarial Training to Composite Semantic Perturbations (Paper)
    • Lei Hsiung, Yun-Yun Tsai, Pin-Yu Chen, Tsung-Yi Ho
    • The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), 2023
  6. Test-time Defense against Adversarial Attacks: Detection and Reconstruction of Adversarial Examples via Masked Autoencoder (Paper)
    • Yun-Yun Tsai, Ju-Chin Chao, Albert Wen, Zhaoyuan Yang, Chengzhi Mao, Tapan Shah, Junfeng Yang
    • The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), AdvML Workshop, 2023
  7. CARBEN: Composite Adversarial Robustness Benchmark (Paper)
    • Lei Hsiung, Yun-Yun Tsai, Pin-Yu Chen, Tsung-Yi Ho
    • International Joint Conference on Artificial Intelligence (IJCAI), 2022
  8. Generalizing Adversarial Training to Composite Semantic Perturbations (Paper)
    • Yun-Yun Tsai, Lei Hsiung, Pin-Yu Chen, Tsung-Yi Ho
    • International Conference on Machine Learning (ICML), AdvML Workshop, 2021
  9. Voice2Series: Reprogramming Acoustic Models for Time Series Classification (Paper)
    • Chao-Han Huck Yang, Yun-Yun Tsai, Pin-Yu Chen
    • International Conference on Machine Learning (ICML), 2021
  10. Transfer Learning without Knowing, Reprogramming Black-box Machine Learning Model with Scarce Data and Limited Resources (Paper) (Video) (Slides) (Code)
    • Yun-Yun Tsai, Pin-Yu Chen, Tsung-Yi Ho
    • International Conference on Machine Learning (ICML), 2020
  11. CloudLeak: Large-Scale Deep Learning Models Stealing Through Adversarial Examples (Paper) (Code)
    • Honggang Yu, Kaichen Yang, Teng Zhang, Yun-Yun Tsai, Tsung-Yi Ho, Yier Jin
    • Network and Distributed System Security Symposium (NDSS), 2020
  12. Adversarial Machine Learning for Social Good: Reprogramming Black-box Machine Learning Model
    • Yun-Yun Tsai, Pin-Yu Chen, Tsung-Yi Ho
    • Neural Information Processing Systems (NeurIPS) NewInML Workshop, 2019
  13. Vehicle Sequence Reordering with Cooperative Adaptive Cruise Control (Paper)
    • Ta-Wei Huang, Yun-Yun Tsai, Chung-Wei Lin, Tsung-Yi Ho
    • Design, Automation & Test in Europe Conference & Exhibition (DATE), 2019

§ Patents

  1. Pin-Yu Chen, Yun-Yun Tsai, Sijia Liu, Chia-Yu Chen, I-Hsin Chung, Tsung-Yi Ho. ”Transfer Learning With Machine Learning Systems”, U.S. Patent Application No: 17/029506, Application Date: Sept. 23, 2020.

Professional Experience

§ Research & Working Experience

§ Honors, Awards, and Grants


Nov. 2023 Awarded the 2023 NeurIPS's scholar award
June 2023 Awarded the 2023 CVPR's scholar award
Sep. 2021 PhD Dean’s Fellowship, Fu Foundation School of Engineering and Applied Sciences, Columbia University.
Aug.12.2020 Best Presenter, Blackhat Award Forum in CyberSec, Taiwan, 2020.

§ Invited Speech


Dec.5.2020 I was invited to give a talk about my ICML'20 paper at Taiwanese Association of Aritificial Intelligence (TAAI) 2020.
Aug.12.2020 I was invited to give a talk about "CloudLeak: DNN Model Extractions from Commercial MLaaS Platforms" at CyberSec 2020 Blackhat Awarded Forum, Taipei.

§ Service

Paper Review CVPR 2023, NeurIPS 2023, IEEE Access, KDD 2021, ICLR 2021, AAAI 2021, ICPAI 2020
Teaching Assistant Engineering Software-as-a-Service, Fundamental of Formal Language, Very-Large-Scale Integration

Extra Activites

    Besides my academic research passion, I am also a versatile musician and have participated in many extracurricular music activities. With over 20 years of performing and training in playing the piano and viola, I was selected as the 1st chair of viola in Tsing Hua Symphony Orchestra for 4 years in college and performed more than 20 concerts on public stages such as National Concert Hall. I also won the NTHU concerto competition in 2017 and was honored to be the piano concerto soloist at the annual concert of NTHU Orchestra.

    § To learn more about my music activities,   click here.

    § To see my video of piano concerto recital,   click here.