Holiday
DELTA台達102 TnT5T6
若論及人工智慧(artificial intelligence, AI)在當代多媒體與文化創意產業之各種技術應用,音樂資訊檢索(music information retrieval, MIR)為其中不可或缺的一環。自2000年國際音樂資訊檢索學會(ISMIR)成立之後,經過二十餘年 的發展,音樂資訊檢索已經成為一個專門而系統化的領域,涵蓋處理各種格式之音樂資料如音訊、樂譜、音樂數位介面等之技 術、對於音樂資料中的音樂資訊如音高、節奏等之辨識技術,自動產生符號與音訊的音樂生成技術,乃至於這些技術關心的使 用者、文化和倫理議題,皆屬音樂資訊檢索的研究問題。換言之,音樂資訊檢索是一門跨領域技術,其技術價值往往建立在與 音樂相關的各種跨領域問題上。 本課程為資訊工程技術導向,將討論許多訊號處理(signal processing)與機器學習(machine learning)技術在音樂資訊 檢索領域的用法。課程內容分為四大部分:1)聲音訊號處理與資料建模技術、2) 自動音樂辨識(recognition)與理解 (understanding)、3)音樂合成(synthesis)與生成(generation)、4) 音樂資訊檢索技術之應用。本課程以課堂講授、作 業、期末專題等方式進行,課程目標為養成團隊研發音樂人工智慧產品之技術力。期末專題製作將強調技術與跨領
Course keywords: Music information retrieval, artificial intelligence, sound and music computing, signal processing, machine learning Part 1: Introduction - The history and the presence of MIR - Fundamentals of signal processing - Fundamentals of machine learning - Fundamentals of information retrieval - Fundamentals of music theory - Fundamentals of psychoacoustics Part 2: Music recognition and understanding - Pitch detection - Chord and tonality analysis - Onset detection - Audio-to-symbolic music transcription - Audio-to-symbolic music synchronization - Tempo estimation and beat tracking - Voice/stream/melody separation - Discovery of repeated patterns - Musical structure analysis - Music classification and tagging Part 3: Sound synthesis and music generation - Sound morphing and synthesis - Music source separation - Sound and voice generation - Rule-based music generation - Harmonization - Counterpoint - style imitation and transfer - Performance generation - Music generation, copyright, and plagiarism Part 4: human-centered MIR with applications - MIR evaluation and reproducibility - Recommendation system - Music and motion - Music performance analysis - MIR and human-computer interaction - MIR and new multimedia arts - Computational musicology Term project: All students taking this course will team up to accomplish a term project, which MUST be related to music. If you want to explore other areas like speech or soundscape, please incorporate them into music application. Interdisciplinary collaboration is highly encouraged. 評分方式:作業 (60%) 期末分組報告 (Term project) (40%)。本課程開放學生使用AI。 參考書: Meinard Müller, Fundamentals of music processing: Audio, analysis, algorithms, applications. Springer, 2015. Mark Gotham, Kyle Gullings, Chelsey Hamm, Bryn Hughes, Brian Jarvis, Megan Lavengood, and John Peterson, Open Music Theory. VIVA Open Publishing, 2023.
MON | TUE | WED | THU | FRI | |
08:00108:50 | |||||
09:00209:50 | |||||
10:10311:00 | |||||
11:10412:00 | |||||
12:10n13:00 | |||||
13:20514:10 | |||||
14:20615:10 | |||||
15:30716:20 | |||||
16:30817:20 | |||||
17:30918:20 | |||||
18:30a19:20 | |||||
19:30b20:20 | |||||
20:30c21:20 |
Average GPA 3.35
Std. Deviation 1.35
本課程為 16 週課程。
-
-
-