The Codec Avatars Lab brings together a world-class and cross-discipline R&D team with the shared goal of developing the next generation of AR and VR experiences. In our Lab, we are revolutionizing the way we interact with each other in the era of the metaverse - in the form of a pioneering technology called codec avatars. These avatars achieve unprecedented realism, yet integrating these avatars into sensing systems that allow live-drivability and controllability remains challenging.A talented Research Engineer is now sought after to own the algorithmic exploration and integration of live-drivable avatars, connecting them to state-of-the-art foundational AI models. This work involves handling of various techniques in generative modeling, understanding and improving complexity and efficiency limitations, and full-stack latency reduction for live systems. An ideal candidate should be passionate in these areas, while contributing to other research engineering topics around codec avatars; strong scientific coding, documentation, and interpersonal communication skills are essential.Join us on the adventure of a lifetime as we make science fiction real and change the world!

• Design, implement and analyze complex live systems that combine various machine learning techniques and connect state of the art generative models with neural renderers
• Collaborate with fellow staff members to develop research ideas, experiments, sensor prototypes, and processing pipelines to advance the SOTA in realistic telepresence
• Deliver readable, maintainable and scalable code in support of these activities
• Analyze and improve latency of large-scale machine learning models such as diffusion models or LLMs

• Currently has, or is in the process of obtaining a Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta.
• Experience with languages and frameworks such as Python, C++, and PyTorch.

• Familiarity with select topics related to realistic representations of human
• Understanding of the full machine vision pipeline from hardware to algorithms
• Demonstrated track record of publication in top conferences such as CVPR, ECCV, ICLR, NeurIPS, and SIGGRAPH
• Experience developing an end-to-end ML pipeline that includes data preprocessing, model development, deployment, and optimization
• Experience with generative models such as Large Language Models and Diffusion Models