Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Deep Visual Odometry
Investigate the usability of a learned visual odometry pipeline for quadrotor flight.
Classical VIO pipelines use geometric information to infer the ego-motion of the camera and couple this information with measurements from the IMU. While these pipelines have shown very good performance in controlled, structured environments, their performance decreases when applied in low-texture or dynamic environments or when applied to high-speed motion.
Recent works propose the usage of data-driven approaches for camera ego-motion estimation. This exploratory work investigates the usability of a learned odometry pipeline for quadrotor flight.
Classical VIO pipelines use geometric information to infer the ego-motion of the camera and couple this information with measurements from the IMU. While these pipelines have shown very good performance in controlled, structured environments, their performance decreases when applied in low-texture or dynamic environments or when applied to high-speed motion.
Recent works propose the usage of data-driven approaches for camera ego-motion estimation. This exploratory work investigates the usability of a learned odometry pipeline for quadrotor flight.
Implement a data-driven visual-inertial odometry pipeline. Test (in simulation, maybe real world) if quadrotor flight is possible with the learned pose estimate.
Implement a data-driven visual-inertial odometry pipeline. Test (in simulation, maybe real world) if quadrotor flight is possible with the learned pose estimate.