Research
My primary research interests lie at the intersection of machine (deep) learning, computer vision, and cyber-physical systems, with applications spanning navigation, positioning, mapping, object detection, and semantic segmentation, among other areas. I broadly categorize these areas as Machine Intelligence and Vision problems. Additionally, I am also interested in interdisciplinary research that integrates these core domains with diverse fields such as disaster response and prevention, environmental monitoring, health sciences, assistive technologies, and smart cities. These interdisciplinary efforts aim to address real-world challenges while leveraging advances in AI and data science.
Within these research areas, my work can be broadly categorized into two main domains: 1) fundamental AI and data science, and 2) interdisciplinary applications. In many cases, the focus is on developing AI, particularly based on deep learning techniques, to address interdisciplinary challenges.
Fundamental AI and Data Science:- Deep Learning for Computer Vision: Developing advanced deep learning techniques for applications like object detection and segmentation. Example papers: RSE 2025
- Innovative AI Learning Strategies: Exploring methods to enhance AI model performance in challenging scenarios such as with incremental/continual learning, self-supervised learning, multi-task learning, few-shot learning, domain adaptation, curriculum learning, etc. Example papers: ICRA 2019, NN 2022
- Model Compression: Investigating approaches like knowledge distillation and quantization to improve deep learning model efficiency for edge computing applications (e.g., IoT, robots, UAV). Example papers: ICCV 2019
- Simultaneous Localization and Mapping (SLAM) and Odometry Estimation: Advancing navigation and mapping technologies using embedded, intelligent, multi-modal sensor data in GPS-denied environment. Example papers: RAL 2020, TRO 2021
- Sensor Fusion: Integrating data from multiple sensors to enhance the accuracy, reliability, and robustness of systems in applications such as robotics, autonomous vehicles, remote sensing, etc. Example papers: SenSys 2020, AAAI 2021
- Distributed and Federated Learning: Building collaborative deep learning systems across decentralized data sources. Example papers: SECON 2021
- Application of Foundational Models: Leveraging vision-based, vision-language, or geospatial foundational models for specialized downstream tasks.
- Disaster Response and Prevention: Developing AI-driven solutions for earthquake and flood detection, impact assessment, emergency response tools, among others. Example papers: GRSL 2024, CPS-ER 2022
- Environmental Monitoring: Assessing environmental impacts, such as the footprints of mining activities, using deep learning-based remote sensing, GIS, and other applications. Example papers: RSE 2025, JEM 2024
- Assistive Technology: Designing AI tools for navigation assistance and scene understanding (using VLM) for visually impaired individuals, and others. Example papers: UIC 2014
- Health Science and Smart Cities: Exploring AI applications to improve public health and urban living standards.
Please refer to publication page for more examples of papers.
Current Research Projects
Please see Monash Research Page for complete list of the projects. In the past, I also contributed to Oxford’s NIST-IPSER and ACE-OPS.
Global Mining Watch
The aim of the project is to perform remote sensing, geo-spatial and socio-environmental analysis of global mining footprints, supporting social impact and risk assessment of mining sites around the world. The project will utilise cutting-edge computer vision and deep-learning approaches, GIS analysis of global data and new approaches to web-based mapping and visualisation of socio-ecological environmental challenges. The project under the ‘Global Mining Watch’ umbrella will be funded from multiple funding bodies, including Google Research Scholar Award, Ford Foundation, and joint initiative between Monash University Indonesia and the University of Queensland.
Funding: Multiple grants from Google Research Scholar Award (PI), Ford Foundation (Co-PI), and UQ-MI Research Collaboration (Co-PI)
Jobs: Multiple positions available for Post-Doc/Research Associate/Research Assistant! (Application closed)
Publications: RSE 2025 [Q1, IF=11.1], JEM 2024 [Q1, IF=8], JED 2023 [Q1, IF=2.3]
Completed Research Projects
Intelligent Remote Sensing for Sustainable Flood Risk Management and Policy
This research seeks to develop an intelligent remote sensing system that can segment and map the temporal dynamics of flooding on a daily basis using AI-based remote sensing approaches to help assess flood impacts and design more sustainable flood management policies. Our study is situated in Indonesia’s upper Citarum river basin where land use change has led to the occurrence of annual flooding events.
Funding: Monash Indonesia Seed Innovation Grant (PI) & MDFI
Jobs: Multiple positions available for Research Associate/Research Assistant! (Application closed)
Publications: IEEE GSRL 2024 [Q1, IF=4], AQUA [Q2, IF=2.1]
Fintech For Social Impact
The project aims to help a fintech institution improve its risk assessment of individuals it might not usually lend to. This project combines expertise in Data Science/Machine Learning, Psychology, Human-Computer Interaction and Economics/Finance.
Funding: Monash Indonesia Seed Innovation Grant, Monash University Australia - Action Lab, and a fintech startup company (Co-PI)
Jobs: Multiple positions available for Research Associate/Research Assistant! (Application closed)
Publications: In preparation.
Regulating Sexual VAWs in Metaverse: An Interdisciplinary Diagnosis
The aim of the project is to legitimise the regulation of sexual violence against women in metaverse by providing various use-based justifications, to explore various Human-Computer Interaction (HCI) in the context of metaverse which has the potential to fertilise VAWs, and to explore and map aspects that may contribute to the development of an ecosystem of trust in metaverse, including by analysing and modelling social media data using machine learning.
Funding: Meta/Facebook (Co-PI), PI is from Monash Malaysia
Jobs: Multiple positions available for Research Assistant! (Application closed)
Publications: In preparation.