Meta AR/VR Job | Research Scientist, Embodied AI
Job（岗位）: Research Scientist, Embodied AI
Type（岗位类型）: Artificial Intelligence | Research
Citys（岗位城市）: Redmond, WA
Date（发布日期）: Before 2021-12-14
Summary（岗位介绍）
At Facebook Reality Labs Research, our goal is to explore, innovate, and design novel interfaces and hardware for virtual, augmented, and mixed reality experiences. We are driving research towards a vision of an always-on augmented reality device that can enable high-quality contextually relevant interactions across a range of complex, dynamic, real-world tasks in natural environments; to achieve this goal, our team draws on methods and knowledge from artificial intelligence, machine learning, computer vision, and human–computer interaction. We are looking for a skilled and motivated researcher with expertise in embodied artificial intelligence (AI), autonomous vehicles, robotics, human–robot interaction, or related fields to join our team. More broadly, the chosen candidate will work with a diverse and highly interdisciplinary team of researchers and engineers and will have access to cutting edge technology, resources, and testing facilities.
In this position, you will work with an interdisciplinary team of domain experts in embodied artificial intelligence (AI), human–computer interaction, computer vision, cognitive and perceptual science, and sensing and tracking on problems that contribute to creating human-centered contextual interactive systems. The position will involve leveraging concepts from embodied AI, autonomous vehicles, and robotics to define and lead research in developing end-to-end policies that leverage diverse multimodal sensors and data sources—including dense 3D reconstructions, egocentric video, audio, gaze, and other signals from wrist-wearable inputs—to predict contextually relevant information that will enhance interactions with future augmented-reality devices. These models will leverage large-scale real-world data sets and the scale of Facebook machine-learning infrastructure, and will be deployed into AR/VR prototypes to uncover research questions on the path to the next era of human-centered computing.
Qualifications（岗位要求）
Experience holding a faculty, industry, or government researcher position
PhD degree in computer science, computer vision, machine learning, artificial intelligence, or related technical field
Demonstrated track record in defining and leading research in embodied AI, autonomous vehicles, autonomous driving, robotics, or human–robot interaction, including developing E2E perception-to-action pipelines, multimodal sensor fusion for scene understanding, deep learning for perception, multi-agent simulation, or related areas
5+ years of experience in at least one deep-learning software library (e.g., PyTorch, Caffe2, TensorFlow, Keras, Chainer). This experience should include formulation, training, and evaluation of new algorithms and writing reusable Python modules
5+ years of experience developing end-to-end ML pipelines in at least three of the following areas: dataset preprocessing, model development and evaluation, software integration, and real-time deployment in embedded systems
Interpersonal experience: cross-group and cross-culture collaboration
Must obtain work authorization in country of employment at the time of hire, and maintain ongoing work authorization during employment
Description（岗位职责）
Develop and execute a cutting-edge research program with interdisciplinary collaborators aimed at developing representations, contextual models, and E2E policy pipelines from multimodal data sources, including 3D reconstructions
Develop tasks, data-collection strategies, modeling approaches, and evaluation criteria to deliver on research program objectives
Work collaboratively with other research scientists to develop novel solutions and models in service of contextualized AI for augmented reality
Mentor MS/PhD interns and postdocs and collaborate with external academic groups to advance our research goals
Additional Requirements（额外要求）
Expertise in sequential decision making methods, including reinforcement learning, dynamic programming, optimal control, and planning
Experience in geometric computer vision, including tracking, 3D reconstruction, localization, object detection, and scene understanding
Experience leading a team of researchers toward executing on a complex technical goal in a cross-functional setting
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, NeurIPS, ECCV/ICCV, ICCP, ICML, 3DV, BMVC, or SIGGRAPH
Familiarity with ideas in representation learning, few-shot learning, and multimodal machine learning
