About the Project
The goal of this project is to capture visual information from real-world environments, and is further reconstructed and converted into editable geometric models with semantic contexts.
Scene understanding may involve information capture and analysis at geometric and semantic levels. The former focus on extraction of geometric entities/primitives from a scene, as well as interactions between them. The latter is mainly to learn the dense semantic labels for each of the geometric primitive obtained from a scene. Properly understanding a scene is an important prerequisite for richer industrial applications, including autonomous system, navigation, mapping, and localisation. The ability to understand a scene depicted in a set of static images with other multi-sensory information has been an essential computer vision problem in practice. However, this level of understanding is rather inadequate since the real-world scene is often dynamic and noisy where unknown objects might be moving independently, as well as visual properties like illumination and texture might change by time.
The aim of this project is to develop deep learning based techniques that (1) enable real-time estimation of both geometric and semantic information from a real-world scene; (2) create 3D editable contents using geometric and semantic information we obtained from a scene. In particular, the outcome of the project should comprise real-time 3D reconstruction for a real-world scene including a series of real-world challenges e.g. dynamic objects, illumination changes, large textureless regions etc. In addition, a set of high quality labels should be achieved for such a raw 3D model, e.g. semantic labels and information about the shape and pose of objects and layouts of the real-world scene. So that the raw model is then turned into a proper representation that can be further edited by average user using visual interactive environment.
The successful candidate will work closely with experts from IAAPS, as well as external collaborators from Imperial College London, Ryerson University and Technical University of Munich. At the end of this PhD project, the candidate will have acquired skills/techniques to plan and undertake independent research, and the candidate will be equipped to follow a variety of different postgraduate career paths.
The successful candidate is expected to have good knowledge of (1) active learning methodology, visual SLAM, as well as popular robotics simulator such as Gazebo; (2) popular large deep learning systems and their maths foundation; (3) coding skills in C/C++ and Python; (4) oral and academic writing skills.
The successful candidate should also have (or expect to achieve) a minimum of a UK Honours degree at 2.1 or above (or the equivalent qualification) in Computer Science, Electronic Engineering or Mechanical Engineering. A master's level qualification or publication in a good venue would also be advantageous.
Non-UK applicants will also be required to have met the English language entry requirements of the University of Bath.
Enquiries and applications:
Formal applications should be made via the University of Bath’s online application form for a PhD in Computer Science.
More information about applying for a PhD at Bath may be found on our website.
Anticipated start date: 4 October 2021.
 Binbin Xu, Wenbin Li, Dimos Tzoumanikas, Michael Bloesch, Andrew J Davison, Stefan Leutenegger. MID-Fusion: Octree-based Object-Level Multi-Instance Dynamic SLAM. International Conference on Robotics and Automation, ICRA 2019.
 Sajad Saeedi, Eduardo da Costa Carvalho, Wenbin Li, Dimos Tzoumanikas, Stefan Leutenegger, Paul H J Kelly, Andrew J Davison. Characterizing Visual Localization and Mapping Datasets. International Conference on Robotics and Automation, ICRA 2019.
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.