My long-term research focuses on harnessing human-centric information to develop AI systems that serve people.
- Multimodal Representation Learning (e.g., VQA, Text-to-image, Multimodal Pre-training Strategies, Multimedia Recommendation, Video Captioning)
- Natural Language Processing (e.g., LLMs, Language Model Training, Zero-shot NLU, Self-iterating LLMs)
- Human values (e.g., Ethics)