This research project focuses on studying different methods for high resolution 3D reconstruction of objects in the near field.
<!-- ## URLLC upperbounds and solution -->
Using wireless signals for sensing and imaging opens up the plethora of applications in which traditional sensing modalities such as camera's or LiDAR's fail. When scenes are occluded, though optical signals maybe scattered away or unable to penetrate these occlusions, wireless signals (and in our case, mmWave) are still able to go through, allowing us to still 'see' in these less than optimal scenarios. Unfortunately, using mmWave signals does not come for free. We must overcome the different ways in which mmWave signals interact with reflectors and the low resolution as compared to optical signals. Thus, in this area of research we aim to study different ways of overcoming these challenges using a combination of machine learning techniques and signal processing.
Using wireless signals for sensing and imaging opens up the plethora of applications in which traditional sensing modalities such as camera's or LiDAR's fail.
When scenes are occluded, though optical signals maybe scattered away or unable to penetrate these occlusions, wireless signals (and in our case, mmWave) are still able to go through, allowing us to still 'see' in these less than optimal scenarios.
Unfortunately, using mmWave signals does not come for free.
We must overcome the different ways in which mmWave signals interact with reflectors and the low resolution as compared to optical signals.
Thus, in this area of research we aim to study different ways of overcoming these challenges using a combination of machine learning techniques and signal processing, often coupled with physical systems such as robotic arms.
Masters projects are available for EPFL students, interested students should have a strong background in signal processing; a background in robotics or hands on work is a plus.
Samah Hussein Youssef
2 months ago