Research
I'm broadly interested in computer vision, specifically generative models, vision-and-language multimodalities, and robustness & interpretability. I'm also excited to explore other interesting topics.
|
|
Diversify Your Vision Datasets with Automatic Diffusion-Based Augmentation
Lisa Dunlap,
Alyssa Umino,
Han Zhang,
Jiezhi Yang,
Joseph E. Gonzalez,
Trevor Darrell
NeurIPS 2023
website /
arXiv /
code
We introduce ALIA, a method which utilizes large vision and language models to automatically generate natural language descriptions of a dataset's domains and augment the training data via language-guided image editing.
|
|
Using Language to Extend to Unseen Domains
Lisa Dunlap,
Clara Mohri,
Han Zhang,
Devin Guillory,
Trevor Darrell,
Joseph E. Gonzalez,
Aditi Raghunathan,
Anna Rohrbach
ICLR 2023 (Spotlight)
website /
arXiv /
code /
blog
We propose LADS, a method that learns a transformation of the image embeddings from the training domain to each unseen test domain guided by language, while preserving task relevant information.
|
|
Delayed and Indirect Impacts of Link Recommendations
Han Zhang,
Shangen Lu,
Yixin Wang,
Mihaela Curmei
FAccT 2023
arXiv /
code
We find that link recommendations have surprising delayed and indirect effects on the structural properties of networks through adapting a simulation-based approach and an explicit dynamic formation model.
|
|
Applied Research Engineer Intern, NVIDIA
June 2024 - September 2024
Developed and released Commercial VILA, a Vision-Language Model trained on commercial images only.
Designed autolabeling workflows for 3D Vision-Language Models development on industrial applications.
|
|
Research Intern, Tencent AI Lab
June 2023 - Feb 2024
Researched on 3D-aware human body generation using diffusion models, ControlNet, and autoencoders.
|
|
Engineering Summer Analyst, Goldman Sachs
June 2022 - August 2022
Developed and modularized a calculator for a loan product, GS Select, to automatically calculate a key financial metric, risk-weighted asset, on a daily frequency.
|
|
Software Engineer Intern, Bytedance
March 2021 - August 2021
Optimized Tiktok's Android Package size and cold startup time, through developing
optimization passes based on Facebook's open-source Android bytecode optimizer ReDex.
|
Misc
At UC Berkeley, I was a member of UPE (CS Honor Society), PBK (Academic Honor Society), and AWE (Association of Women in EE&CS). From 2020 to 2021, I served as the Director of the International Affairs Department in ASUC Senator Rex Zhang's office.
In my spare time, I enjoy watching movies and dining out. My top 3 movies are Before Sunrise, Blade Runner, and Infernal Affairs (无间道).
|
Website template from Jon Barron. Thanks for stopping by :)
Last updated: Sep 22, 2024
|
|