Research Intern, Spatial Intelligence

What you'll do

• Research, design, and implement methods for generative AI for videos and reconstruction.
• Develop techniques for simulation, video understanding, policy learning, and reinforcement learning.
• Work on Vision-Language Models (VLMs) within the spatial intelligence lab.
• Publish original research in top-tier venues at the end of the internship.
• Write and execute code in Python, C++, and CUDA for rapid prototyping.
• Utilize machine learning frameworks like PyTorch to build and train models.
• Manage code and version control using systems like Git.

What we're looking for

• Must currently be at least a second-year Ph.D. student in Computer Science, Engineering, or Mathematics.
• Excellent knowledge of theory and practice of computer vision methods and deep learning.
• Excellent programming skills in rapid prototyping environments such as Python.
• Proficiency in C++ and parallel programming, such as CUDA.
• Knowledge of common machine learning frameworks, such as PyTorch.
• Outstanding research track record.
• Very good publication record in computer vision with a minimum of 2 papers at top venues (CVPR, ICCV, ECCV, NeurIPS, ICML, ICLR).
• Excellent communications skills and good knowledge of version control systems like git.