Edge AI for Computer Vision Training in Kuala Lumpur, Malaysia

This course provides hands-on training in using Edge AI for computer vision applications, focusing on deploying vision-based AI models on resource-constrained edge devices. Participants will explore real-time image and video processing, object detection, face recognition, and model optimization for hardware like Raspberry Pi, Jetson Nano, and mobile devices.

Foundations of Edge AI and Computer Vision

Introduction to Edge AI and Computer Vision

• What is Edge AI?
• Why computer vision at the edge matters
• Benefits: low latency, privacy, offline processing

Edge AI Hardware and Software Stack

• Overview of devices: Raspberry Pi, Jetson Nano, Coral Dev Board, mobile
• AI toolkits: TensorFlow Lite, OpenVINO, PyTorch Mobile
• Vision libraries: OpenCV, MediaPipe, YOLO models

Computer Vision Basics for the Edge

• Image classification vs object detection vs segmentation
• Key concepts: bounding boxes, inference time, accuracy
• Example walkthrough: classifying handwritten digits on a Pi

Model Optimization and Vision Use Cases

Optimizing AI Models for Edge Deployment

• Quantization, pruning, model simplification
• Model conversion for TFLite and ONNX
• Testing for performance and accuracy trade-offs

Real-Time Object Detection on Edge Devices

• YOLOv5 and SSD models for detection
• Running detection models with live video input
• Comparing accuracy vs. speed on edge hardware

Use Case Walkthroughs

• Smart surveillance: person detection
• Retail: shelf monitoring
• Manufacturing: fault detection
• Medical: monitoring patient posture

Deployment, Integration, and Final Project

Building Vision Applications on the Edge

• Creating pipelines with camera, processing, display
• Managing input/output in real-time
• Logging, alerts, and remote monitoring

Face Recognition and People Tracking (Optional Advanced)

• Running face detection models
• Adding simple tracking or recognition with OpenCV
• Privacy and ethical considerations

Wrap-Up and Next Steps

• Review tools, platforms, and datasets
• Best practices for edge deployment
• Career paths and future trends in Edge AI
• Final Q&A and feedback