###Vision-based Navigation for Soft Endoluminal Robots
Project ID: 2228bd1038 (You will need this ID for your application)
Research Theme: Artificial Intelligence and Robotics
UCL Lead department: Computer Science
Lead Supervisor: Sophia Bano
Project Summary:
Why this research is important? Endoluminal interventions (e.g. endoscopy, otoscopy) are minimally invasive procedures for disease diagnosis and treatment. Limited field-of-view, constrained maneuverability of the endoscope, scene texture paucity, and poor video feedback during such interventions hinder the complete inspection of the endoluminal environment, leading to missed diagnosis and treatment. To improve dexterity and precision during these interventions, soft robots are being developed which are yet to be equipped with artificial intelligence capacities to allow real-time accurate navigation. Unlike real-world mobile robot navigation, where a multitude of sensors (multiple cameras, LiDAR, GPS, IMU, etc) are available, a monocular miniature camera is the only sensing source making the endoluminal robot navigation problem non-trivial. The problem will focus on developing vision-based robot localization and mapping techniques, enabling complete inspection of the endoluminal environment.
Who you will be working with? The Surgical Robot Vision group at UCL WEISS specializes in Surgical Artificial Intelligence (AI) and Surgical Robotics and is working in close collaboration with surgeons for developing novel AI-guided technologies for achieving clinical translations. Alongside 1-to-1 weekly meetings, the team meets on a regular basis for progress updates and guidance to support technical advancement.
What you will be doing? The project will involve the design and development of vision-based deep learning techniques for robot pose and camera depth estimation followed by localization and 3D reconstruction of the endoluminal environment using solely the sensing information provided by the onboard monocular camera. The developed navigation system will be deployed in real time and will be paired with the robot control system through team collaboration. An additional area of research includes endoluminal scene understanding alongside mapping and reconstruction.
Who we are looking for? Undergraduate in computer science, biomedical engineering, mechatronics or a related subject with knowledge of machine learning, computer vision or/and artificial intelligence.