We are seeking a highly motivated postdoctoral researcher to join our team at TU Darmstadt, Germany, working on efficient multimodal (vision and language) models. The goal of this research is to develop highly efficient multimodal models w.r.t. computational budget with an optimal trade-off w.r.t. performance.

The position is for one year, starting as soon as possible, with potential options for extension.

• Conduct research on efficient multimodal AI models, including developing novel architectures, training methods, and optimization techniques.
• Collaborate with an industry partner to ensure that the research has a real-world impact.
• Publish high-quality research papers in top venues such as CVPR, NeurIPS, and ICCV.
• Present research findings at conferences and workshops.

• PhD in Computer Science, Machine Learning, or a related field.
• Research experience in efficient deep models and/or multimodal AI.
• A track record of publishing in top venues such as CVPR, NeurIPS, ICCV, ICLR, ACL, and AAAI.
• Excellent communication and interpersonal skills.

Please apply via our Online Application Form and use “Postdoc-efficient-MMAI” as the “reference number”.