Meta AR/VR Job | Research Scientist, 3D Reconstruction and Eye Tracking | Oculus
Type（岗位类型）: Engineering | Machine Learning, Research
Citys（岗位城市）: Redmond, WA
At Reality Labs Research (RLR), our goal is to make great consumer virtual, augmented, and mixed reality experiences that ship in five to ten years. Come work alongside a world leading team of interdisciplinary scientists and engineers to create the technology that makes VR and AR pervasive and universal. Our research focuses on developing comprehensive solutions to complex eye tracking problems, including both algorithms and hardware for AR and VR systems. We are currently seeking researchers with experience in 3D Reconstruction and Neural Rendering to create highly detailed and accurate models of the eye and face. The 3D models you build will enable our team to better understand the human eye and build better eye trackers.
PhD degree in Computer Science, Machine Learning, or a related field.
5+ years experience in geometric computer vision including object pose estimation, 3D deep learning, 3D face and body reconstruction.
1+ years of experience with Neural Rendering or Neural Implicit methods such as NERF.
3+ years of programming experience in Python or C++ and hands-on experience with frameworks such as PyTorch.
Must obtain work authorization in the country of employment at the time of hire and maintain ongoing work authorization during employment.
Research and develop novel 3D Vision and Machine Learning based methods in the area of 3D Eye and Face Reconstruction.
Work directly with capture hardware to obtain the best-possible data for these reconstructions.
Scale the data collection and reconstruction systems to capture the full diversity of human eye appearance, shape and behavior.
Validate the accuracy of the 3D reconstruction system by comparison to other scanning systems and through end-to-end metrics using these reconstructions for applications.
Collaborate with other research scientists, hardware and software engineers to develop innovative 3D reconstruction techniques for eye tracking use-cases.
Experience with 3D reconstruction of the human eye or other 3D geometrical work with eyes such as model-based eye tracking.
Experience modeling hard-to-capture geometry such as thin structures, subsurface scattering, reflective and refractive surfaces.
Demonstrated experience using simulations from 3D reconstructions to drive design of novel vision systems, or to generate machine learning training sets.
Proven track record in publishing papers in computer vision, graphics and/or machine learning conferences including CVPR, ICCV, ICRA, ISMAR, ECCV, NeurIPS, AAAI, ICLR, TVCG, SIGGRAPH.