Worcester Polytechnic Institute-Object Identification for Surgery Robot and Autonomous vehicles

Keywords:​ Artificial Intelligence, Object Identification, Deep learning

Recommendation: ​This is a project designed for students who have background in computer science or Robotics Engineering. It is recommended for any students who target to gain research experience by means of a Cutting-edge research topic about Object detection and classification which is used for Navigation, Pick and Place and additional robotics activities.

Introduction:

Surgery robots have been developed to assist surgeons in performing challenging tasks accurately and safely which are difficult even for expert surgeons like Vitreoretinal surgery. Recent years have provided great progress in object detection mainly due to machine learning methods that became practical and efficient. Also new data representation and models contributed to this task. Object detection algorithms, activated for robotics, are expected to detect and classify all instances of an object type (when those exist). They should be detected even if there are variations of position, orientation, scale, partial occlusion and environment variations as intensity. Object detection is the key to other machine vision functions such as building 3D scenes, getting additional information of the object (like face details) and tracking its motion using video successive frames. Robotic application, as mentioned, navigation and pick-place, may require more elaborate information from images. In this case, additional image capturing channels may be used. Self-navigating robots, like autonomous vehicles, use multi camera setup, each facing a different direction. The computer vision system employs data fusion during or post the object detection algorithms.

The goal is to help students understand the way of conducting research in computer science and Robotics Engineering, like how to search state-of-the-art material and read papers in the right way, how to generate and optimize a novel intuition to be an excellent research and how to design experiments to validate the research. Above methodology of research would be practiced under instruction using one Cutting-edge topic. By learning search engines knowledge and reading state-of-the-art papers, students are encouraged to generate novel algorithms or architectures to improve the performance or optimize the stat-of-art research. Meanwhile, it is helpful to understand how Robots recognize objects effectively. Students are recommended to write a report including the background of the object recognition in robotic vision, the design of new algorithm and expected experiment performance.