Heekyung (Anne) Lee

Hi! I am Heekyung(Anne) Lee. I am an Undergraduate Student in Computer Science at POSTECH, South Korea. I am also an undergraduate researcher at Berkeley AI Research (BAIR), fortunate to be supervised under Prof. Trevor Darrell’s group.

I am interested in studying Vision and Language. My current research focuses on applications of vision-language models, such as image captioning, visual question answering, and multimodal reasoning.

My recent research interests are on Multi-Modal Language Models.

• Hallucination in Vision Language Models
  ∘ What is hallucination in Vision Language Models, and what are effective training recipes to mitigate it? How can we address hallucinations caused by misalignment between image patches and text tokens?
• Real-World Evaluation of Vision Language Models
  ∘ What is the gap between the performance of VLMs on benchmark datasets and in real-world tasks? How do Vision Language Models respond to complex tasks, and how should we analyze their responses to guide further improvements?

Feel free to check out my publications and projects.

.

And also, lets be Strava friends! (Follow me here) I love running XD.