External Research Grants

(This is additional funding to SMU with a project extension.)

This project, led by A/Prof Iris Rawtaer (Sengkang General Hospital) aims to utilise multimodal sensor networks for early detection of cognitive decline. Under this project, the SKH team will oversee the project operations, screening recruitment, psychometric evaluation, data analysis, data interpretation, reporting and answer of clinical research hypotheses. The SMU team will collaborate with SKH to provide technical expertise for this study by ensuring safe implementation and maintenance of the sensors in the homes of the participants, provide the sensor obtained data to the clinical team and apply artificial intelligence methods for predictive modelling.

This research collaboration aims to examine the interrelated dimensions of governance, national identity, and belonging among Singaporeans citizens aged 60 and above. The research ultimately aims to contribute to a deeper understanding of how government performance and political narratives influence citizen engagement and well-being in Singapore, thereby providing valuable recommendations for enhancing social cohesion and democratic participation.

(This is additional funding to SMU for Stage 1b of the project.)

State-of-the-art visual perception models in autonomous vehicles (AV) fail in the physical world when meeting adversarially designed physical objects/environmental conditions. The main reason is that they are trained with discretely-sampled samples and can hardly cover all possibilities in the real world. Although effective, existing physical attacks consider one or two physical factors and cannot simulate dynamic entities (e.g., moving cars or persons, street structures) and environment factors (e.g., weather variation and light variation) jointly. Meanwhile, most defence methods like denoising or adversarial training (AT) mainly rely on single-view or single-modal information, neglecting the multi-view cameras and different modality sensors on the AV, which contain rich complementary information. The above challenges in both attacks and defenses are caused by the lack of a continuous and unified scene representation for the AV scenarios. Motivated by the above limitations, this project firstly aims to develop a unified AV scene representation based on the neural implicit representation to generate realistic new scenes. With this representation, we will develop extensive physical attacks, multi-view & multi-modal defenses, as well as a more complete evaluation framework. Specifically, the project will build a unified physical attack framework against AV perception models, which can adversarially optimize the physical-related parameters and generate more threatening examples that could happen in the real world. Furthermore, the project will build the multi-view and multi-modal defensive methods including a data reconstruction framework to reconstruct clean inputs and a novel ‘adversarial training’ method, i.e., adversarial repairing that enhances the robustness of the deep models with guidance of collected adversarial scenes. Finally, a robust-oriented explainable method will be developed to understand the behaviors of visual perception models under physical adversarial attacks and robustness enhancement.

Retail banks use real-time monitoring, machine learning, security checks, and rule-based systems to detect fraud by spotting deviations from normal customer behaviour. Quantum computing could dramatically boost these systems by processing vast datasets faster, enhancing pattern recognition, anomaly detection, and encryption. This project will build on prior quantum innovations and partnerships to identify and validate quantum algorithms that outperform existing fraud detection methods in accuracy and speed. It will also estimate required quantum resources, protect intellectual property, and train staff in quantum techniques, aiming for commercial advantage and contributing valuable knowledge to the financial industry’s fight against fraud.

Singapore’s openness - as a hub for the global exchange of goods, ideas, and people – brings many opportunities but also leaves its local firms and workers uniquely susceptible to shocks originating outside the country. With increasing geopolitical and economic uncertainty, understanding the mechanisms through which global shocks (such as trade wars and supply chain disruptions) propagate across firms' linkages becomes particularly relevant for the economy. The proposed research program seeks to establish datasets on firm-to-firm supply linkages for the Singapore economy, and to identify the mechanisms via which external and internal shocks transmit via these linkages to affect the firm-level performances, such as productivity, labor demand, and innovation activities. Once established, the local economy’s firm-to-firm dataset will open up a new landscape for academic research as well as policy analysis, given Singapore’s uniquely high degrees of participation in the global value chain, heavy reliance on multinational enterprises, and strong presence of service sectors.

This project aims to equip Android application analysers with AI-driven input fuzzer and output interpreter to make them usable products even for those without domain expertise.

Large Language Models (LLMs) are increasingly applied to sectors such as healthcare, finance, software development and autonomous driving. However, their complex and interconnected supply chains—including data pipelines, inference frameworks, software dependencies, and deployment infrastructures—introduce significant security, reliability, and ethical risks. These complexities amplify vulnerabilities and increase the potential for system-wide failures, necessitating a holistic, system-level approach to risk identification and mitigation. This project aims to systematically address these risks through three key objectives: (1) developing comprehensive risk assessment methodologies for the entire LLM supply chain, (2) designing LLM-specific cyber insurance products to mitigate potential losses, and (3) collaborating with industry partners to ensure practical adoption and real-world impact. By tackling these challenges, the project will enhance the trustworthiness, security, and sustainability of LLM deployment across critical domains.

(This is additional funding to SMU with a project extension.) 

Funded by the Ministry of Law, this initiative, known as the Singapore Global Restructuring Initiative, is housed under the SMU Centre for Commercial Law in Asia, and it seeks to demonstrate international thought leadership on restructuring and insolvency.

This exploratory study investigates the impact of Singapore Management University’s (SMU) interdisciplinary Core Curriculum in developing lifelong learners. It focuses on three interrelated questions: (1) How do educators in SMU’s Core Curriculum approach interdisciplinarity, and (if relevant) lifelong learning, in their curricular design, course delivery and assessment?; (2) To what extent does SMU’s interdisciplinary Core Curriculum develop lifelong learning among its graduates?; and (3) What are the best practices in interdisciplinary education that develop lifelong learning that can be shared with Singapore IHLs?

This project tackles critical challenges in the development and deployment of Multimodal Large Foundation Models (MLFMs), which are capable of understanding and generating content across text, image, video, and audio modalities. While current MLFMs exhibit impressive performance, their trustworthiness, accuracy, and high operational costs limit their accessibility—especially for smaller research groups and organizations.

To address these gaps, the project focuses on two key innovations: (1) developing fine-grained "super alignment" techniques to reduce hallucinations and ensure model outputs align with human values, and (2) creating a scalable, low-cost multi-agent framework composed of smaller specialized models (8-15 billion parameters) that work collaboratively on complex tasks. These innovations will be powered by Reinforcement Learning with Human Feedback (RLHF) and Reinforcement Learning with AI Feedback (RLAIF), enabling continuous refinement and adaptation.

The research will be validated through real-world applications such as video generation and multimodal chatbots, demonstrating both practical utility and cross-domain adaptability. Ultimately, this work aims to democratize access to advanced AI, supporting Singapore’s strategic goal of building inclusive, trustworthy, and globally competitive AI capabilities.

This research/project is supported by the National Research Foundation, Singapore under its National Large Language Models Funding Initiative (AISG Award No: AISG-NMLP-2024-002). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore.