Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Real-time, learning based 3D gaze estimation on an embedded system mounted on a motorcycle
As a motorcyclist it is important to keep your eyes on the road at all times. However,
arguably more important is where you look while riding. In particular when manoeuvring
through curves the correct gaze location of the rider is critical for a safe ride.
As a motorcyclist it is important to keep your eyes on the road at all times. However,
arguably more important is where you look while riding. In particular when manoeuvring
through curves the correct gaze location of the rider is critical for a safe ride. Novice
riders typically look only a couple of meters in front of the rider while expert riders look
towards the end of the curve. Similarly checking whether the rider has seen hazardous
objects such as pedestrians and vehicles can be achieved with 3D gaze estimation.
As part of a joint master thesis between the Computer Vision Lab and Aegis Rider AG an
ETH Pioneer Fellowship startup, you will develop a real-time, learning based method to
detect the 3D gaze position of a motorcycle rider. The end goal of this thesis is to know
exactly where the rider is looking in 3D space.
**Requirements**: C++ & Python skills, having worked with ROS1/2, knowledge on deeplearning
frameworks such as Pytorch or Tensorflow including embedded optimisations via
TensorRT
As a motorcyclist it is important to keep your eyes on the road at all times. However, arguably more important is where you look while riding. In particular when manoeuvring through curves the correct gaze location of the rider is critical for a safe ride. Novice riders typically look only a couple of meters in front of the rider while expert riders look towards the end of the curve. Similarly checking whether the rider has seen hazardous objects such as pedestrians and vehicles can be achieved with 3D gaze estimation. As part of a joint master thesis between the Computer Vision Lab and Aegis Rider AG an ETH Pioneer Fellowship startup, you will develop a real-time, learning based method to detect the 3D gaze position of a motorcycle rider. The end goal of this thesis is to know exactly where the rider is looking in 3D space.
**Requirements**: C++ & Python skills, having worked with ROS1/2, knowledge on deeplearning frameworks such as Pytorch or Tensorflow including embedded optimisations via TensorRT