Geometric Vision Engineer – Perception and Localization

Vimaan is seeking a skilled and enthusiastic Computer Vision and Robotics Engineer to play a pivotal role in the development and implementation of advanced geometric vision, localization and SLAM algorithms for our state-of-the-art warehouse inventory products. This role involves working on the development and implementation of Python-based modules and pipelines to be deployed on both server and edge-based platforms and requires thinking out-of-the-box to build modules that are both optimal and efficient.

Geometric Vision Engineer – Perception and Localization Responsibilities

• Design, development and implementation of on-board computer vision algorithms, using both classical and modern methods, meant for real-time robot localization in an indoor warehouse setting.

• Implement sensor fusion techniques to integrate data from multiple sensors such as LIDAR, Cameras, IMU and ToF to enhance the accuracy and robustness of vehicle navigation.

• Implement marker and tag detection algorithms using classical and modern computer vision techniques for real-time localization.

• Contribute to geometric vision algorithms involving 3D computer vision tasks such as dense/sparse 3D reconstruction, Multiview pose estimation, RGBD fusion, 3D object detection and 3D dimensioning.

• Contribute to the product architecture, make decisions on sensor selections, and deploy solutions in various customer sites and gauge system efficiency on a day-today basis.
  

• Characterize and build metrics on the effectiveness of the computer vision algorithms in tackling the problem at hand.
  

• The role will include assessment of the best-in-class technologies, choice of the right technologies to implement for the Vimaan use case, implementation of the technology into our real-life solutions at customer facilities and continued support and improvement of the technology in the field.

Geometric Vision Engineer – Perception and Localization Qualifications

– Master’s degree in Robotics, Computer Science, Navigation and Control Systems or related field.

– Strong Python programming skills. Responsible for conducting thorough testing and debugging of Python code, ensuring the reliability and robustness of pipeline implementations.

– Strong foundation in physics and robotics systems – including ROS, 3D simulations and sensor fusion based navigation algorithms.

– Proven experience in developing Navigation and Localization algorithms for autonomous vehicles, robotics, drones, maritime vehicles or similar application

– Hands-on experience in computer vision (classical and modern), visual odometry and SLAM projects, solving real-world problems involving vision, navigation and localization tasks.

– Deep expertise in Kalman filtering, Extended Kalman filtering, and other estimation techniques and knowledge of sensor fusion and data integration techniques

– Deep mathematical foundations with knowledge of 3D Multi-view Geometry, Advanced Linear Algebra, Numerical Optimization, etc.

– Deep insights into depth cameras and 3D data characteristics.

– Ownership of the development and implementation of modules for computer vision and packaging software for efficiency.

– Ability to understand problem statements and implement and test new ideas under supervision.

– Maintain comprehensive documentation of code, algorithms, and pipelines.

– Capability to work collaboratively in a team and communicate effectively.

– Ability to work in a fast paced, high pressure startup environment and adapt to rapidly changing requirements.

– High proficiency in C++ and Python, with allegiance to writing well designed and modular code.

Role Level:

Mid-Level. 1-2 years of work experience.