Meta AR/VR Job | Research Engineer Intern, Multimodal Data Alignment/3D Reconstruction
Job（岗位）: Research Engineer Intern, Multimodal Data Alignment/3D Reconstruction
Type（岗位类型）: 3D Software Engineering | Computer Vision, Engineering, Research
Citys（岗位城市）: Boston, MA
Date（发布日期）: Before 2021-12-14
Summary（岗位介绍）
Due to occlusion or limited input imagery, 3D reconstruction techniques and semantic segmentations produce incomplete 3D point clouds and map features. Completing missing data and obtaining spatial relations between objects extracted from various data sources are vital for various AR / VR applications. We have multiple projects in the team which focus on addressing these issues. These projects involve improving the point clouds produced by structure-from-motion (SfM) algorithms and features extracted from these imagery by utilizing spatial and geometric priors and using techniques like crossview matching. We will combine multiple data sources like street-level imagery, high resolution oblique aerial imagery, and features like buildings and roads from Facebook maps to improve 3D point clouds and derive useful features from these datasets. We will review existing work on this domain and explore both geometry-based and machine learning techniques. The goal of these projects is to build an efficient and scalable solution to the problems based on research ideas.
Qualifications（岗位要求）
Currently has, or is in the process of obtaining a PhD in Computer Vision or a related field.
2+ years experience in 3D Computer vision, geometry and Machine Learning.
Experience coding in Python and C++.
Experience in complex and high volume 3D Data coming from different sources.
Must obtain work authorization in the country of employment at the time of hire and maintain ongoing work authorization during employment.
Description（岗位职责）
Literature review on existing methods related to 3D point clouds and mesh completion, crossview matching and alignment and semantic segmentation.
Perform research on scene understanding by developing prototypes based on state of the art algorithms and approaches, extending those approaches to address the specific needs of AR/VR platforms.
Conduct analyses of the algorithms and prototypes being researched and developed.
Communicate plans, progress, and results to the team and the broader organization.
Write software to implement, evaluate, and visualize the algorithms and prototypes being researched and developed.
Collaborate with other engineers and researchers across various disciplines.
Additional Requirements（额外要求）
Ability to communicate complex research in a clear, precise, and actionable manner.
A strong interest in theoretical and empirical research and for answering challenging questions with research.
Prior experience working with frameworks like PyTorch3D and SfM.
Prior research experience and, ideally, prior publication(s) in some of the top computer vision venues (CVPR, ECCV, ICCV, BMVC, etc.).
Intent to return to degree-program after the completion of the internship.
