Dr Salem Ameen
School of Science, Engineering & Environment
Current positions
Lecturer
Biography
Dr Salem Ameen is a Lecturer in Artificial Intelligence, Robotics and Automation at the º£½ÇÂÒÂ×. His research focuses on efficient, robust and practically deployable artificial intelligence for robotics, perception, and real-world decision-making.
He completed his PhD at the º£½ÇÂÒÂ× in 2017, where his research specialised in optimising deep learning networks using multi-armed bandit methods. Building on this foundation, his current interests span modern deep learning, computer vision, multimodal AI, robotics, and intelligent autonomous systems.
His recent research direction includes transformer-based learning, vision-language and multimodal reasoning, robot perception in challenging environments, embodied AI, and physically grounded common-sense reasoning. He is particularly interested in intelligent systems that can move beyond surface prediction toward richer representations of structure, space, and real-world dynamics.
Dr Ameen has supervised projects in computer vision, deep learning, reinforcement learning, robotics, and machine learning. He welcomes postgraduate researchers interested in contemporary AI methods with strong real-world relevance, especially in robotics, perception, multimodal intelligence, healthcare AI, and autonomous systems.
Areas of Research
Efficient Deep Learning and Model Optimisation: Developing scalable and resource-aware deep learning methods, including optimisation and efficient model design for practical deployment.
Transformers and Modern Deep Learning Architectures: Investigating transformer-based models and related architectures for vision, multimodal learning, and intelligent autonomous systems.
Computer Vision and Robot Perception: Building robust perception methods for robotics and automation, including RGB-D sensing, scene understanding, and perception in challenging environments such as mirrors, glass, reflective, and transparent surfaces.
Vision-Language and Multimodal AI: Exploring models that integrate visual, spatial, and language information for reasoning, perception, and decision support. Recent CVPR work shows strong momentum in this area, including spatial reasoning VLMs and multimodal perception-language systems.
Embodied AI and World Models: Studying predictive and representation-based learning methods for planning, control, and real-world understanding, including compact latent-space world models.
Common-Sense and Physically Grounded Reasoning: Investigating AI systems that can reason about spatial relationships, likely object behaviour, and real-world structure in support of robotics and intelligent decision-making. Recent work on visual common sense and 3D vision-language reasoning makes this a timely research direction.
AI for Healthcare and Automation: Applying machine learning and intelligent decision-making methods to healthcare, diagnostics, automation, and industrial systems.
Areas of Supervision
PhD Topics
Deep Learning
Efficient Deep Learning
Transformer Models
Vision Transformers
Multimodal AI
Vision-Language Models
Spatial Reasoning in AI
Common-Sense Reasoning for Intelligent Systems
Embodied AI
World Models and Predictive Representation Learning
Computer Vision
Robot Perception and Sensing
RGB-D Perception in Challenging Environments
Autonomous Systems
Human-Robot Interaction
Machine Learning for Healthcare
AI for Automation and Intelligent Systems
Level 7:
Artificial Intelligence
Mobile Robotics
Interactive Visualization
Automation and Robotics
Mechatronics
Level 5:
Numerical Analysis
Computing Laboratory (Numerical Methods and Simulation)
Level 4:
Probability
Mathematics and Computing
Level 3:
Mathematics 1
Mathematics 2
Qualifications
-
Machine learning
2013 - 2017 -
Computer Science and Engineering
2007 - 2009