Master Thesis Projects in Multimodal Foundation Models

12.11.2024, Nichtwissenschaftliches Personal

If you're passionate about multimodal foundation models and eager to explore the integration of vision, language, and other modalities in a unified framework, we invite you to join our cutting-edge research team. Dive into the future of AI by developing models that can understand and interact across multiple forms of data, unlocking new levels of intelligence and versatility.

We are seeking talented students to mentor, aiming to produce high-quality publications with practical applications. Our team has a strong track record, with publications in top international conferences such as NeurIPS, ICLR, ACL, EMNLP, NAACL, AAAI, UAI, and CIKM. Equipped with state-of-the-art Nvidia A100 GPUs, our research is at the forefront of multimodal machine learning, capable of handling extensive datasets and complex algorithms.

As part of this thesis project, you will have the opportunity to collaborate with a diverse group of PhDs, postdoctoral researchers, and professors from prestigious institutions like TUM, LMU, and others worldwide. This experience will allow you to develop advanced research skills and delve into the exciting field of World Models. Join us on this groundbreaking research journey and make a significant impact in the world of AI.

Topic Description:
Concrete topics in Multimodal Foundation Models

Requirements:
· Study in computer science, electrical engineering, physics, or mathematics;
· Good understanding of machine learning and deep learning;
· Good understanding of transformer as a plus;
· Good programming skills in Python and at least one deep learning framework;
· Independent working

Additional Information:
Start: from now on
Length of the work: 6-9 months

Kontakt: cognitive.yunpu@gmail.com, xingcheng.zhou@tum.de