Dr Sankaran Iyer
PhD: Computer Science, UNSW 2023
MCompSc: UNSW 1994
BE (Hons): Electrical and Electronics Engineering from Birla Institute of Technology and Science Pilani (India)
Dr Sankaran Iyer obtained his PhD from UNSW Sydney in 2023, where his research focused on vertebral compression fracture detection using a novel 3D localisation framework that combined deep reinforcement learning and imitation learning. His work explored supervised, weakly supervised and semi-supervised learning approaches for medical image analysis, localisation and segmentation.
He also completed a Master's degree in Computer Science at UNSW in 1994, with research focused on Latin character detection using artificial neural networks.
Dr Iyer has over 30 years of industry experience spanning real-time embedded systems, intelligent networks and operations support systems. Prior to returning to academia, he worked at Nokia and Alcatel-Lucent, where he voluntarily retired in 2016 as a Senior Project Manager.
He has collaborated with researchers from the Biological, Earth and Environmental Sciences (BEES) group at UNSW on projects involving house dust mite and pest detection systems, as well as an Android-based wildlife species detection application developed as part of the Bushfire Recovery program.
Currently, Dr Iyer is a Senior Research Associate at UNSW working in collaboration with the Black Dog Institute on AI-based suicide detection and prevention research. His work focuses on behaviour analysis in public environments such as railway stations, bridges, parks and shopping centres using pedestrian detection, multi-object tracking, pose estimation and anomaly detection techniques.
His broader research interests span intelligent perception systems, computer vision and embodied AI, including:
- Object detection and multi-object tracking
- Behaviour analysis and anomaly detection
- Medical image analysis, localisation and segmentation
- Intelligent monitoring and surveillance systems
- Reinforcement learning and learning-based decision making
- Thermal and low-light computer vision
- Drone detection and tracking
- AI for safety-critical environments
- Embodied AI and autonomous robotics
Dr Iyer is actively exploring embodied AI and autonomous robotics, with particular interests in intelligent inspection, autonomous monitoring and resource-constrained robotic systems operating in real-world environments. Recent work includes the development of a low-cost embodied AI surveillance robot integrating embedded sensing, IMU-assisted navigation, environmental scanning and deep learning-based person detection.
His current interests include multi-modal robotic perception, edge AI, continual learning and intelligent autonomous systems that combine perception, reasoning and action in the physical world.
Further information on Dr Iyer's current activities in AI, Computer Vision, Embodied AI and Autonomous Robotics, including research projects and consulting interests, is available at his personal website:
- Publications
- Media
- Grants
- Awards
- Research Activities
- Engagement
- Teaching and Supervision
My research activities focus on Artificial Intelligence, Computer Vision and Embodied AI, particularly in the areas of:
Embodied AI and Autonomous Robotics
- Autonomous mobile robots
- Multi-modal robotic perception
- Sensor-guided navigation and exploration
- Intelligent inspection and monitoring systems
- Edge AI for robotics
- Learning-based autonomous behaviour
Computer Vision and Intelligent Perception
- Object detection and multi-object tracking
- Behaviour analysis and anomaly detection
- Surveillance analytics and intelligent monitoring systems
- Human activity recognition
- Thermal and low-light computer vision
- Omnidirectional and multi-camera vision systems
Intelligent Monitoring and Safety Systems
- AI for safety-critical environments
- Intelligent transport systems
- Drone detection and tracking
- Intruder detection and security monitoring
- Smart parking systems
- Retail analytics and inventory monitoring
Medical Imaging and AI for Healthcare
- Medical image localisation
- Medical image segmentation
- Boundary-aware medical image analysis
- Weakly supervised and semi-supervised learning
Currently, I work as a Senior Research Associate at UNSW in collaboration with the Black Dog Institute, focusing on AI-driven behaviour analysis for suicide detection and prevention in public environments such as railway stations, bridges, parks and shopping centres. This work involves pedestrian detection and tracking, pose estimation, spatio-temporal behaviour analysis and anomaly detection using deep learning techniques.
My current robotics activities are independent of my Senior Research Associate role and focus on practical embodied AI systems for intelligent monitoring, autonomous inspection and low-cost robotic platforms for research and education.
My work combines research and practical system development, with a strong emphasis on deployable AI systems operating in complex real-world environments.
My Research Supervision
I am currently co-supervising 2 PhD students and guiding a Master of Information Science student. Additionally, I assist other students with coding and model building in PyTorch, TensorFlow, and other deep learning platforms.