Meta AR/VR Job | Computer Vision Engineer – XR Input ML Systems
Job（岗位）: Computer Vision Engineer – XR Input ML Systems
Citys（岗位城市）: Burlingame, CA
Facebook Reality Labs is helping people around the world come together and connect through world-class virtual and augmented reality (VR/AR) hardware and software. With global departments dedicated to VR/AR research, computer vision, haptics, social interaction, and more, we are committed to driving the state of the art forward through relentless innovation.
VR/AR’s potential to change the world is immense—and we’re just getting started. Our XR Tech team explores, develops, and delivers new cutting-edge technologies that serve as the foundation of current and future VR/AR products. From mixed reality and human interaction to natural inputs and beyond, XR Tech is focused on taking new technologies from early concept to product while iterating, prototyping, and realizing the human value and new experiences they open up.
The XR Input ML Systems team is responsible for bringing ML-based hand tracking capabilities to our current and upcoming AR/VR devices. Our team bridges the gap between research and production, and is responsible for optimizing our hand tracking models within the constraints of each hardware device.
As a CV specialist working on the ML Systems team, you will be expected to solve end-to-end applied CV & ML problems associated with improving our hand tracking models. Problems range from improving data collection and synthetic data usage, optimizing ML models given various hardware constraints, identifying where our models fail and reducing these gaps, and optimizing our models to incorporate new poses and gestures that are important to the end-user experience.
Bachelor’s degree in relevant computer science field
5+ years of experience in at scale data engineering, data analysis/visualization, telemetry and operations
5+ years of industry experience in C/C++ and Python
5+ years of experience in using applied CV/ML to solve user-facing problems
Invent, analyze, design, develop and optimize real time VR and AR computer vision algorithms in areas related to 3D reconstruction, SLAM, object detection and tracking, sensor fusion and camera calibration
Collaborate in a team environment across multiple scientific and engineering disciplines, making the architectural trade-offs required to rapidly deliver software solutions
Develop software for a variety of consumer devices, including novel sensing systems, tracking systems, imaging systems and haptic devices
Produce clean readable code, debug complex problems that span systems, prioritize ruthlessly and get things done with a high level of efficiency
Learn constantly, dive into new areas with unfamiliar technologies, and embrace the ambiguity of AR and VR problem solving
Background in Computer Vision – pose estimation, action recognition, SLAM, object detection, Natural Language Processing – automatic speech recognition, statistical text decoding, or Robotics – state estimation
PhD in a related field of Computer Vision, Computer Graphics, Natural Language Processing or other AI/Machine Learning related field
Deep knowledge of modern machine learning techniques, Bayesian filtering, information theory, and 3D geometry