Multimodal Deep Learning for Resilient and Robust Remote Sensing Semantic Segmentation (Dual PhD Programme)

This PhD project, titled 'Multimodal Deep Learning for Resilient and Robust Remote Sensing Semantic Segmentation,' offers a unique opportunity to advance the state of the art in semantic segmentation for remote sensing data. The project is part of a prestigious 4-year Dual PhD degree programme between the National Tsing Hua University (NTHU) in Taiwan and the University of Liverpool in England, granting successful candidates two PhD awards from world-leading institutions. Students will benefit from international research experience, access to large-scale national research facilities, and the chance to build a global network of contacts. The research focuses on developing a next-generation AI-based framework for fine-grained semantic segmentation using multimodal remote sensing data, such as optical, hyperspectral, radar, and thermal imagery. The project addresses critical challenges including class imbalance and missing data, leveraging cutting-edge Vision-Language Foundation Models and Multimodal Large Language Models. This pioneering approach will be the first to jointly tackle multimodal fusion, missing data, and class imbalance within a unified semantic segmentation framework, making the technology more applicable to real-world scenarios where such issues are common. Outcomes from this project have significant potential across multiple application domains, directly contributing to the UN Sustainable Development Goal (SDG) 9. The methods developed will support data-driven decision-making for infrastructure resilience, sustainable urban growth, and resource-efficient industrial development. Authorities and industry stakeholders will be able to identify infrastructure deterioration, monitor pollution and emissions, assess construction impacts, and track progress towards sustainable land-use practices. By advancing AI for remote sensing, the project will strengthen the innovation ecosystem supporting sustainable infrastructure planning and environmental management. The doctoral programme is structured so that students spend the first two years at NTHU under the supervision of Professor Shang-Hong Lai, followed by two years at the University of Liverpool under the supervision of Dr Keiller Nogueira. Upon completion, students will be awarded two PhD degrees, one from each institution. Both universities have agreed to waive tuition fees for the duration of the project and provide a maintenance stipend: TWD 15,233/month for two years in Taiwan, and the UKRI Studentship rate (£20,780/year for 2025-26, rising with inflation) for two years in Liverpool. Additional funding is available through a Research Training Support Grant for consumables and conference attendance. Scholarships are available for outstanding international students. Applicants should have a strong academic background in computer science, artificial intelligence, machine learning, or related fields, with a bachelor's or master's degree required. Experience with deep learning, computer vision, or remote sensing is highly desirable. The opportunity is open to both home (UK) and international students, and English language proficiency requirements apply as per University of Liverpool guidelines. To apply, candidates should complete the University of Liverpool online postgraduate research application form for a PhD in Computer Science, including the project title and reference number NTHU009. Dr Keiller Nogueira and Professor Shang-Hong Lai should be listed as proposed supervisors. For further information or questions, contact Dr Keiller Nogueira at [email protected]. This project represents a unique chance to contribute to impactful research at the intersection of AI, remote sensing, and environmental science, while gaining international experience and dual qualifications from two leading universities.