.png?width=22&height=21&name=Artificial%20Intelligence%20(AI).png){kind=small}
Computer Vision (CV) focuses on processing and understanding images, while Vision Intelligence (VI) extends this by applying Machine Learning (ML) to achieve human-like interpretation of visual data. Autonomous Systems (AS) integrate such capabilities with sensor inputs to make decisions independently. Together, they enable applications such as self-driving vehicles, drones, and urban safety solutions. Although short courses on these topics exist, they are often too general or lack domain-specific focus, motivating a proposed hands-on course centered on defense and urban use cases like drone tracking, public safety monitoring, and workplace safety enforcement. Drone tracking—highlighted by recent conflicts—requires detecting and continuously monitoring drones. Camera-based solutions offer a low-cost option under line-of-sight conditions. In urban settings like Singapore, video analytics has enhanced policing efficiency, with Police Cameras (PolCams) helping solve over 5,000 cases since 2012. In high-risk workplaces such as construction sites, systems like SafeSite VA have been trialed to detect unsafe practices (e.g., missing helmets), improving compliance and reducing incidents.
Learning Outcomes
1. Explain concepts of image processing, computer vision, and autonomous systems
2. Develop basic programs in Python that can be run locally as well as in the cloud
3. Train a basic computer vision model for detecting and localizing a fixed object in a room
4. Understand basic model quality metrics and apply them to improve the training process
5. Develop a functional computer vision and autonomous systems prototypes
Day 1
Morning
1. Introduction to Image Processing and Computer Vision
2. Introduction to Autonomous Systems
3. Introduction to Python
4. Introduction to OpenCV with Python
Afternoon
1. Introduction to Jupyter Notebook and Google Colab
2. Introduction to YOLO
3. Training a YOLOv8 model
4. Drone tracking using YOLOv8 and Google Colab
Day 2
Morning
1. Human landmarks recognition using MediaPipe
2. Detecting public unrest using human landmarks recognition
3. Safety violation detection using YOLOv8 and MediaPipe
Afternoon
1. Demo of trainee specified CV and AS application
2. End of course feedback
.png?width=50&name=Artificial%20Intelligence%20(AI).png)
-1.png?width=360&name=Server%20(4)-1.png)
/AI%20(6)-1-1.png?width=360&name=AI%20(6)-1-1.png)
/AI%20(8)-1.png?width=360&name=AI%20(8)-1.png)
-1.png?width=360&name=Automation%20RPA%20(2)-1.png)
/AI%20(5)-1.png?width=360&name=AI%20(5)-1.png)
