終端人工智慧之實體技術與應用

臺北科技大學 115-1 學期課程，課號 366624。

• 學期：115-1
• 授課教師：洪子晏、范書愷
• 學分：3
• 選修
• 上課時段：週一 2、3、4 節

課程大綱

本課程介紹終端人工智慧之實體技術與應用的核心概念與實務應用，聚焦於智慧系統如何透過感知、決策與行動與實體世界互動。課程內容涵蓋實體 AI 系統的重要基礎技術，包括機器學習感知模型、神經網路訓練、強化學習、模擬環境，以及系統整合等主題。並透過實作練習理解終端人工智慧於機器人、智慧裝置、自主系統等實體場域之實際應用建立對新興智慧實體技術發展趨勢的整體認識。 This course introduces the core concepts and practical applications of PhysicalTechnology and Edge AI, focusing on how intelligent systems interact with the physical world through perception, decision-making, and action. The course covers fundamental technologies of embodied AI systems, including machinelearning-based perception models, neural network training, reinforcement learning, simulation environments, and system integration. Through hands-on exercises, students will gain an understanding of the practical applications of Edge AI in physical domains such as robotics, smart devices, and autonomous systems, while developing a broad perspective on emerging trends in intelligent physical technologies.

課程進度

1. Introduction to Edge AI and the applications 2. Simulation Platforms and Isaac Lab 3. Supervised Learning for Perception 4. Neural Networks and Model Training 5. Visual Perception for Embodied AI 6. Reinforcement Learning Fundamentals 7. Q-Learning and Deep Q-Networks 8. Running and Modifying RL Tasks 9. Integrating Perception and Decision 10. Project Development and Presentation

評量方式

Homework and practice 50% Term Project 50%

教材

【遵守智慧財產權觀念，請使用正版教科書，不得使用非法影印教科書】 使用外文原文書：是 Li, B., et al. Embedded Artificial Intelligence: Principles, Platforms and Practices. Springer, 2024. Sutton, R. S., & Barto, A. G. Reinforcement Learning: An Introduction, 2nd ed. MIT Press, 2018.