Computer Vision Engineer (Perception)

Key Skills

Apache Hadoop, C Programming, C++, Data science techniques, Java Programming, Machine learning techniques, MapReduce, Python Programming, PyTorch, R Programming, Scala Programming, SQL, TensorFlow

Job Description

At Path we work on state-of-the-art artificial intelligence, machine learning, computer vision, and sensors to make industrial robots intelligent. Our transformative technology, like our robotic welding system, enables hardware to do more with less human input. We are looking for a Computer Vision Engineer for our Computer Vision Sensor Team, to design and release algorithms into production software and cutting-edge sensor products. The role will focus on fundamentals of vision cameras and 3D sensors, computer vision and graphics, a range of image processing algorithms and high level computer vision topics, convolutional neural networks for visual recognition and both deep learning and classic methods for point cloud applications.
You will report to the Sensor Team Lead and will join a team of dedicated, supportive, and enthusiastic people to
help create the future of manufacturing.
What you will do

• Develop and implement algorithms for high-speed, high-accuracy, versatile 3D
  reconstruction or imaging and stereo vision systems
• Develop novel sensor and robot calibration algorithms, techniques, procedures and
  software and hardware devices
• Develop high performance computer vision programmes with CUDA and GPU
  programming
• Work on scalable sensor solutions that include software, optical, electrical and mechanical
  designs and verifications methods
• Work on scalable calibration solutions that include software, optical, electrical and
  mechanical designs and verifications methods
• Stay abreast of new research and novel findings in the fields of computer vision and
  especially reconstruction of the complex visual world
• Test and evaluate current state-of-art algorithms and techniques of computer vision and
  graphics, especially for 3D metrology, calibration and point cloud applications
• Research important 3D point cloud principles and applications
• Partner with our other engineering teams in artificial intelligence, robotics, and software
  engineering

Who you are

• You have written a PhD dissertation or a MS thesis in the field of computer vision or something strongly related
• You have 3+ years of experience in developing advanced computer vision algorithms or applications, especially in such topics as 3D metrology, fast imaging, multiple-view geometry, epipolar geometry, structure from motion, geometric and topological data analysis, calibration, ray tracing, reflection, object tracking, SLAM or other cutting edge computer vision or robotics research

• You have high-level familiarity with point cloud registration and other applications
• You have good familiarity in conventional neural networks for images
• You are highly proficient in C++ and computer vision libraries
• You are familiar with GPU programming, CUDA and Nvidia Jetson
• You are proficient at calculus, linear algebra, probability and statistics