This course explores cutting-edge techniques for creating efficient machine-learning models to address the growing demand for real-time decision-making and localized processing across diverse application fields, including IoT/robotics/smart manufacturing systems and beyond. Key topics include model compression, pruning, quantization, neural architecture search, knowledge distillation, on-device fine-tuning, transfer learning, application-specific acceleration techniques, etc. Through hands-on projects, students will learn to optimize and adapt deep learning models for resource-constrained devices while maintaining accuracy and performance.
Unlock more information by signing in with your HKUST account.