Push the borders of what’s achievable in real-time Vision-based localisation and mapping by developing new algorithms that can fuse multiple camera and IMU data at speeds upwards of 150 km/h. The candidate will work on the cutting edge of Visually aided localization for aerial vehicles.
Requirements-
Must have:
● 3+ years experience in C++
● Understanding of computer vision concepts like camera models, multi-view geometry, and structure-from motion
● Experience with Feature engineering for vision and IMUs
● Understanding of image processing pipeline
● Implemented algorithms using UKF, EKF, Factor Graphs
● Conceptual/Implementation experience with Vision-based algorithms like ROVIO, VINS-MONO, SVO, DSO, or DTAM
● Understanding of Visual-Place recognition
● Experience with simulation of algorithms in a Linux-based platform
Good to have:
● Code optimisation for embedded devices
● Worked on Nvidia based compute boards
● Working knowledge on control systems, motion planning and/or deep learning
Personality Traits:
● Strong Analytical thinking and problem solving ability
● Ability to work with a distributed and diverse team of researchers, developers, and engineers
● Ready to learn new concepts, stay up-to-date with the state-of-the-art in the field and implement them
● Ability to work within deadlines and dynamic work environment
● Excellent English language communication skills, both written and verbal