Computer Vision / Deep Learning Engineer (m/w/d)

  • ANavS GmbH
  • München, Germany
  • 20/11/2020
Full time Data Science Data Analytics Big Data Data Management Statistics

Job Description

Job advertisement Id FT003 online since 01/2020

You develop computer vision and deep learning approaches for processing camera and LiDAR sensor data to perform visual/ LiDAR odometry and SLAM. The computer vision based pose measurements shall support and be integrated into the ANavS sensor fusion framework for precise position and attitude determination. Alongside the focus on precise positioning, the precise mapping and the interpretation of the environment are in focus as well. You will explore and develop machine learning and state-of-the-art deep learning techniques not only for processing camera and LiDAR data, but also for performing monitoring of sensor inputs, filter states and (intermediate) outputs of the sensor fusion framework.

Your tasks:
  • Development of computer vision and machine learning/ deep learning approaches to perform visual/ LiDAR odometry and SLAM from camera and LiDAR data and to perform analysis and interpretation of the environment
  • Development of machine learning/ deep learning approaches to monitor the sensor fusion inputs, states and outputs
  • Integration of computer vision supporting sensors, such as IMU and wheel odometry
  • Integration of precise GNSS-based pose measurements for high quality mapping
  • Implementation of interfaces between computer vision derived measurements and the sensor fusion framework
  • Multi-sensor calibration incl. camera calibration, multi-sensor spatial and time calibration/ synchronization
  • Contribution to the development of a general computer vision framework
  • Evaluation of developed positioning and mapping solutions
  • Contribution to application-driven customer projects, as well as research-oriented funded projects
  • Bringing developed solutions towards product stage
  • Patenting/ publishing of developed approaches is encouraged
Your profile:
  • Experience in computer vision and machine learning/ deep learning (at least 2 years)
  • Computer science degree (or comparable)
  • Well-founded knowledge in C++ programming
  • Well-founded knowledge of (visual) SLAM techniques
  • Experience in processing camera and LiDAR data for localization and mapping
  • Experience with deep learning frameworks, such as TensorFlow or Caffe
  • Practical experience in designing, developing, training and evaluating deep neural networks that solve computer vision tasks, such as detection, segmentation and tracking on camera or LiDAR data
  • Preferable experience in using ROS, Python and GPU programming (CUDA)
  • High motivation to contribute to the technical development, ability to work independently and willing to adapt to flexible tasks
  • Team player with quick perception, reliability and accuracy
  • Good communication skills in English and German