The aim of this module is to provide students with a practical understanding of deep generative models and their applications in robotics. Students will learn how to implement and apply generative models, such as Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Diffusion Models, to solve robotic perception, control, and decision-making challenges. This module is suitable for students interested in applying generative AI techniques to real-world robotic systems and projects, emphasizing practical skills and project-based learning.
Students completing the module successfully should be able to: A. Understand the core concepts of deep generative models, focusing on their use in robotics, including VAEs, GANs, and Diffusion Models. B. Implement and train generative models using deep learning frameworks, such as PyTorch, to address robotic tasks. C. Analyze and evaluate the performance of generative models in robotic applications using appropriate metrics and visualization techniques. D. Apply generative models to enhance robotic perception, environmental modeling, and control, and collaborate effectively within project teams to solve complex problems.
The teaching philosophy of the module follows very much the philosophy of Syntegrative Education. This means that the teaching delivery pattern, which follows more intensive block teaching, allows more meaningful contributions from industry partners. This philosophy is carried through also in terms of assessment, with a reduction in the use of exams and an increase in coursework, especially problem-based assessments that are project-focused. This module will be delivered through a combination of lectures, group discussions, case studies, hands-on practical exercises, etc. Lectures and group discussions are conducted using the Problem-Based Learning paradigm focusing on student-centred learning, where students develop critical thinking and problem-solving skills to address open-ended problems that lack a straightforward solution. This module is taught with an emphasis on student learning through practice and by projects, facilitated by a module leader, and where appropriate, industrial mentors. Students can identify particular areas of learning needs or interests according to the available project(s). They will conduct independent research to gather information and resources to better define the problem. Progress towards the learning outcomes will be facilitated and monitored, where students are guided to progressively address the given problem through tasks. Independent learning will form an important aspect of the educational activities in this module. Assessed by a project, students shall gain practical experience in undertaking independent study and research on industry-focused real-world problems.