Research Scientist Intern, 3D World Modeling, Robotics and Generative AI (PhD)

Meta; Sunnyvale, Burlingame, Redmond (CA, CA, WA); On-site; USD 60,000 - 120,000; Full time

Job summary

Meta Reality Labs Research Team seeks exceptional researchers to advance egocentric machine perception and digital twins for AR/VR devices. The internship focuses on cutting-edge computer vision and ML techniques across 3D scenes and world models.

You will collaborate with researchers on egocentric world models, multimodal data, and end-to-end model training from data collection to evaluation, with multiple start dates throughout the year.

Qualifications

• PhD in computer vision, ML, or related field and hands-on experience in 3D perception.
• Strong publication record at CVPR/ICCV/SIGGRAPH or NeurIPS is a plus.
• Proficiency with Python and PyTorch in research projects.

Responsibilities

• Collaborate with researchers to advance egocentric world models and action-conditioned video synthesis.
• Explore multimodal input modalities and 3D data processing.
• Implement 3D computer vision algorithms and end-to-end ML training pipelines.
• Contribute to data curation and evaluation for scientific publications.

Skills

Python

PyTorch

3D computer vision

End-to-end ML training

Unix environment

Education

PhD in computer vision/ML

Tools

Git

Unix

Job description

The Meta Reality Labs Research Team brings together a world-class team of researchers, developers, and engineers to create the future of AR and VR. The Surreal Vision group at RL Research is seeking exceptional Research Scientists to research and help build the egocentric machine perception functionalities and digital twins of 3D scenes that will underpin future VR and contextual-AI-enabled AR devices. The research intern will work on cutting edge research problems to innovate novel computer vision and machine learning techniques.

Collaborate with researchers to advance work across multiple research areas, including:

• Developing novel egocentric world models for action-conditioned video synthesis.
• Evaluating different egocentric input modalities (e.g., hand pose, head pose, eye gaze, body pose).
• Establish multimodal generation models including RGB video, 2D/3D flow, and depth maps.
• Curating and post-processing large-scale datasets to enable effective training.
• Investigating the impact of world models on RL algorithms for policy learning.
• Benchmarking against state-of-the‑art approaches and compiling results for scientific conference submission.

Our internships are twelve (12) to twenty-four (24) weeks long and we have various start dates throughout the year.

Currently has, or is in the process of obtaining a PhD degree in the domain of computer-vision, computer graphics, 3D machine perception or machine learning Knowledge and hands‑on experience working with 3D and projective geometry, in addition to image‑space computer vision Hands‑on experience implementing 3D computer vision algorithms as well as end‑to‑end training of ML models from data collection and design to evaluation Experience working within Python environments such as pytorch Experience working in a Unix environment Must obtain work authorization in the country of employment at the time of hire, and maintain ongoing work authorization during employment Proven track record of achieving significant results as demonstrated by grants, fellowships, patents, as well as first‑authored publications at leading workshops or conferences such as CVPR, ECCV/ICCV, SIGGRAPH/SIGGRAPH Asia, ICLR and NeurIPS Strong track‑record of published research in the fields of neural reconstruction, inverse rendering, neural rendering, object reconstruction and tracking, generative modelling Strong programming experience using python and pytorch Demonstrated software engineer experience via an internship, work experience, coding competitions, or widely used contributions in open source repositories (e.g. GitHub) Intent to return to a degree‑program after the completion of the internship Experience working and communicating cross functionally in a team environment