Postdoc and PhD opportunities in Computer Vision, Generative AI, and World Models at KAIST KCVL Lab Korea Advanced Institute of Science and Technology in South Korea

KAIST Computer Vision and Learning Lab (KCVL), led by Prof. Tae-Kyun (T-K) Kim, is recruiting postdoctoral researchers through the InnoCORE Fellowship 2026, with additional broader openings for PhD students and postdocs in computer vision and machine learning.

The research themes are centered on Computer Science, Computer Vision, Machine Learning, Generative AI, World Models, and related areas such as Physical AI, video and 3D generation, diffusion and flow models, multimodal foundation models, embodied AI, AI4Science, robotics, pose estimation, object detection and tracking, and reinforcement learning.

For the postdoc call, candidates should have a PhD completed or expected, within 5 years of graduation as of May 1, 2026, and strong publications at top venues such as CVPR, ICCV, NeurIPS, ICML, ICLR, SIGGRAPH, ECCV, PAMI, IJCV, or TIP are preferred. The lab also looks for candidates who can lead research independently and collaborate effectively in a large group.

Funding for the postdoctoral position is about 90M KRW per year through the InnoCORE Fellowship, running through February 2027 with possible extension. The post also mentions substantial compute resources, including about 32× H100 GPUs and additional H200 capacity being secured.

To apply, email your CV and Google Scholar link to [email protected]. The job page states a submission deadline of around April 21, 2026, or until the position is filled. The lab also notes that it is always looking for strong PhD and postdoc candidates in computer vision and machine learning.