Research Scientist Intern, Robust Tracking via 3D Gaussian Splatting (PhD)

The Reality Labs (RL) Research Team at Meta is seeking a Research Scientist Intern for Robust Tracking via 3D Gaussian Splatting. This internship focuses on solving dense correspondence problems in processing input data of human subjects using 3D Gaussian Splatting methods. The goal is to improve robustness in challenging scenarios of longer-term correspondences, exploring theoretical aspects and potentially developing quasi-convex loss functions for more robust tracking.

Research topics include:

• Virtual humans and animatable avatar creation
• Dense motion tracking
• Differentiable rendering via 3D Gaussian Splatting
• Machine learning / AI

The internship duration is 12 to 24 weeks, with various start dates throughout the year. The role involves designing and executing algorithms, investigating novel approaches in body capture, tracking, reconstruction, and generative models for clothed humans and deformable objects. The intern will prototype and analyze experimental systems, collaborate with researchers across disciplines, and communicate research progress and results.

Ideal candidates should have or be pursuing a PhD in Computer Vision, Graphics, Machine Learning, or a related field. Experience with computer vision, machine learning, 3D geometry, and linear algebra is required, along with proficiency in Python and PyTorch. Familiarity with 3D Gaussian Splatting is beneficial.

This internship offers an opportunity to work on cutting-edge AR and VR technologies at Meta, contributing to the future of immersive experiences and social technology.