Thor Research Lab is a research group in Artificial Intelligence and Natural Language Processing, based in Ho Chi Minh City and led by Dr. Tran Khai Thien. Our work centres on Multimodal data for Vietnamese, that is, on models that read text and images together rather than in isolation.

The lab works across Artificial Intelligence, Natural Language Processing, Computer Vision and Multimodal Learning. Within these areas, our current topics include multimodal classification, multimodal question answering, vision-language representation and reasoning, and the construction of datasets and benchmarks for Vietnamese. We are also interested in how pre-trained vision-language models and multimodal large language models can be adapted to Vietnamese data, and in how such models behave when the data they receive is noisy or incomplete.

We welcome students, visiting researchers and collaborators who share an interest in Vietnamese language technology. If our work overlaps with yours, we would be glad to hear from you.