Embedded AI Systems Project
Published:
Project Overview
Embedded AI project for COMP4901D: Embedded AI Systems at HKUST, Spring 2026.
Status: In Progress (Expected completion: May 2026)
Objectives
Developing and deploying AI models on resource-constrained embedded devices, focusing on optimization and real-time performance.
Key Challenges
- Model optimization for embedded systems
- Real-time inference on limited hardware
- Power efficiency and performance trade-offs
Technologies Used
- TensorFlow Lite / PyTorch Mobile
- Embedded platforms (Raspberry Pi, NVIDIA Jetson, etc.)
- C++/Python
- Edge computing frameworks
Course
COMP4901D: Embedded AI Systems, HKUST, Spring 2026
Paper Presentation
As part of the course requirements, I presented a research paper on denoising generative models:
Paper: Back to Basics: Let Denoising Generative Models Denoise
Presentation: View Slides on Overleaf
The presentation introduces the key concepts and contributions of this paper on denoising diffusion models.