人工智慧於衛星系統應用

臺北科技大學 115-1 學期課程，課號 367156。

課程大綱

本課程主要目的在於使用各種深度學習的方法，應用在微波遙測資料上面。主要以案例的方式介紹各種應用，因此除了會說明核心的演算法之外，也會介紹使用的訓練資料集的格式、前處理等內容。課程前段會先回顧基本的遙測影像特性以及深度學習技術，包含CNNs, RNNs, GANs和YOLOs等，常見於遙測資料應用的方法。課程中段會已實際的範例說明上述幾種方法的使用情境與成效。課程最後會要求學生以文獻研讀的方式說明不同深度學習方法在微波遙測上的實例。 The primary objective of this course is to apply various deep learning methods to microwave remote sensing data. It introduces diverse applications primarily through case studies, covering not only core algorithms but also the format of training datasets and preprocessing techniques. The initial part reviews fundamental remote sensing image characteristics and deep learning techniques, including CNNs, RNNs, GANs, and YOLOs—methods commonly applied to remote sensing data. The middle part demonstrates the practical scenarios and effectiveness of these approaches through real-world examples. The final part requires students to conduct literature reviews, presenting case studies of different deep learning methods applied to microwave remote sensing.

課程進度

Week 1, Syllabus Week 2, Fundamental remote sensing image characteristics - 1 Week 3, Fundamental remote sensing image characteristics - 2 Week 4, Review on variant CNNs Week 5, Review on variant GANs & RNNs Week 6, Review on variant YOLOs Week 7, Case study on the GAN: MSTAR dataset generation Week 8, Case study on the RCNN & YOLO: Ship detection using SAR data Week 9, Case study on the U-net and optimzation: SAR & Optical remote sensing images image fusion Week 11, Case study on the Convolutional Learning Machine: AutofocusSAR, Machine learning-based Synthetic Aperture Radar (SAR) autofocus algorithms Week 12, Case stduy on the NN: SAR2SAR, A Semi-Supervised Despeckling Algorithm for SAR Images Week 13, Literature reviews and presenting by students Week 14, Literature reviews and presenting by students Week 15, Literature reviews and presenting by students Week 16, Literature reviews and presenting by students Week 17-18, (Flexible Learning) Final presentation report

評量方式

出席狀況與課程參與(Attendance & Participation): 20% 文獻閱讀與報告(Literature reviews and presenting): 60% 期末文獻閱讀報告(Report): 20%

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：否 相關網路資源