近期演講記錄
2025
- (20250119)Keynote speech at FinNLP (60-min talk)
Title: Transforming Financial Services with AI: Applications, Challenges, and Future Directions
Abstract:
This talk will explore the pivotal role of artificial intelligence in financial institutions and the challenges it must overcome. We will present examples of AI applications in the financial sector, including decision-making AI technologies, such as check image recognition, credit card fraud detection, automated property valuation, and ID card OCR, as well as generative AI solutions, such as legal credit reports, chatbots, knowledge management systems, and ATM image analysis. Additionally, we will delve into critical topics like data governance, model maintenance, and federated learning for cross-bank collaboration. These elements are essential for ensuring the successful implementation of AI technologies in real-world financial scenarios.
本演講將闡述人工智慧在金融機構中所扮演的重要角色及其面臨的挑戰。我們將分享AI在金融業的應用範例,包括決策式AI(如支票影像識別、信用卡詐欺偵測、自動房屋估價、身分證OCR等)和生成式AI(如法人徵信報告、聊天機器人、知識管理、ATM 影像分析等)。此外,我們還將討論資料治理、模型維護和聯邦式學習(用於跨行合作)等議題,這些議題都是決定人工智慧技術能否成功應用於實際金融場景的關鍵因素。
2024
- (20240907)台積電主題演講 (90-min talk)
Title: 金融如何擁抱 AI
- (20240907)國科會「管理一及管理二學門聯合成果發表會」主題演講 (30-min talk)
Title: AI 的應用與未來 - 以金融為例
- (20240827)資策會「AIGO 領袖人才發展產業論壇」演講 (20-min talk)
Title: 賦能平台創新:生成式AI 在金融的應用與推廣
- (20240808)天下雜誌「城市高峰論壇」演講 (20-min talk)
Title: AI 在金融的應用與演進
- (20240313)教育部「AI Cup 2023 年頒獎典禮」主題演講 (50-min talk)
Title: AI 在金融的應用與挑戰
- (20240110)金財通商務科技公司年末策略會議演講(60-min talk)
Title: 金融業導入AI實戰經驗
2023
- (20231211)「AI 與未來投資布局及策略」研討會專題演講 (30-min talk)
主辦單位:衛生福利部
Title: AI 發展與產業趨勢
- (20231103)「2023年臺灣網際網路研討會(TANET)暨全國計算機會議(NCS)」之NCS專題演講 (75-min talk)
主辦單位:TANET-NCS 研討會/政治大學資科系
Title: AI 於金融的應用與展望
- (20231031)「MIC Forum Fall」專題演講 (30-min talk)
主辦單位:資策會 MIC
Title: 金融業導入AI實戰經驗
- (20231021)「同行致遠班」專題演講 (80-min talk)
主辦單位:陽明交大後EMBA「同行致遠班」課程
Title: AI 在金融的應用與展望
- (20231019)頤賢講座專題演講 (100-min talk)
主辦單位:「頤賢講座-臺灣政經社問題與對策」課程
Title: AI 在金融的應用與展望
- (202310005)Ansys Executive Forum 專題演講 (30-min talk)
主辦單位:Ansys at Taiwan
Title: AI/ChatGPT Successful Application Experience
- (20230915)專題演講 (30-min talk)
主辦單位:中國銀保監
Title: 玉山在人工智慧的發展與趨勢
- (20230905)21屆億載會例會專題演講 (60-min talk)
主辦單位:億載會
Title: 人工智慧如何成就金融業 -兼論ChatGPT發展趨勢
- (20230905)專題演講 (100-min talk)
主辦單位:全球人壽
Title: 人工智慧在金融的應用與展望
- (20230726)「金融專業研習會」專題演講 (180-min talk)
主辦單位:法官學院
Title: 人工智慧及金融科技應用
- (20230617)「第 34 屆國際資訊管理學術研討會」Keynote Speech (40-min)
主辦單位:中華民國台灣資訊管理學會、中山大學
Title: 數位轉型與人工智慧在金融的機會與挑戰
Abstract:
受到新冠肺炎疫情影響,數位轉型的步伐與零接觸經濟的推動正在持續加速進行中,同時也造成 AI 的技術被無遠弗屆地應用在金融業。此次演講將說明 AI 在金融業的不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,包含決策式的 AI(例如票據影像辨識、信用卡盜刷偵測、房屋自動估價等)以及生成式的 AI(例如反洗錢、聊天機器人、徵信報告等)。因應 AI 的落地,對應的數位轉型也有多項基礎建設要進行,例如資料的治理、流程的合規、模型的精進、人員的管理等,這些維運面的問題通常不是技術導向,卻是數位轉型及 AI 服務能夠長久運行於金融場景的的最重要關鍵。
- (20230614)「ChatGPT在金融與企業之發展與應用論壇」專題演講 (20-min talk)
主辦單位:台灣大學金融科技研究中心
Title: 生成式 AI 在金融的應用
- (20230530)中央銀行專題演講 (180-min talk)
主辦單位:中央銀行
Title: 人工智慧如何成就金融業-兼論ChatGPT發展趨勢
- (20230428、Fri)Ansys 專題演講 (60-min talk)
主辦單位:Ansys
邀請人:Norman Chang
Title: Opportunities and Challenges of Finance Embracing AI
Abstract:
This talk aims to demonstrate the vital role that AI plays in the financial industry and the challenges it encounters. We will provide examples of AI applications in finance, including AML blacklist detection, bill image recognition, credit card fraud detection, automatic housing valuation, speech recognition and synthesis, and ID card OCR, etc. Furthermore, we will brief address data governance and model management as crucial elements to consider. These maintenance and operational issues are critical factors in determining whether an AI technology can be effectively applied to real-world scenarios and continuously improved in finance.
本次演講旨在說明AI在金融業扮演的重要角色以及面臨的挑戰。我們將分享金融業中AI的實例,如AML黑名單偵測、票據影像辨識、信用卡盜刷偵測、房屋自動估價、語音辨識與合成、身份證OCR等關鍵應用。此外,我們也會探討資料治理、模型管理和精進等方面,這些維運問題是決定某一AI技術能否成功應用於實際場景並不斷改進的關鍵因素。
- (20230324)【第九屆台灣永續趨勢與分析發表會】專題演講 (90-min talk)
主辦單位:致德國際股份有限公司
Title: 科技驅動永續管理 創造金融新未來
- (20230217)【2023 AI 策略治理與案例分析董總班】專題演講 (60-min talk)
主辦單位:台灣人工智慧學校
Title: 金融擁抱AI的機會與挑戰
2022
- (20221209)淡江大學「金融科技理論與實務」系列課程演講
Title: AI在金融領域的應用
- (20221118)台灣人工智慧年會專題演講
Title: 從辨識到維運:AI專案落地的下一步
Abstract: This speech will explain how an AI model can be improved in various directions after the service is launched, and analyzes two practical application examples in E.SUN Bank. At the same time, we will also address the essential elements of a complete AI service to form AI 2.0, and explain how these elements can eventually form the AI ecosystem of the entire company, making the overall AI services more comprehensive and stable. (本次演講將闡述 AI 模型在服務上線後如何在各個方向進行改進,並分析兩個在玉山銀行的實際應用案例。 同時,我們還將探討形成 AI 2.0 所必須具備的完整 AI 服務的基本要素,並解釋這些要素如何最終形成整個公司的 AI 生態系統,使整體 AI 服務更加完善與穩定。)
- (20221108)財政部財政資訊中心專題演講
Title: 人工智慧及金融科技簡介
- (20220725)法官學院講座『111年金融專業研習會(線上學習)』
Title: 人工智慧及金融科技應用
- (20220624)第二屆金融法與金融科技研討會
Title: 玉山金控的AI轉型之路
- (20220524)陽明交通大學資訊工程學系
Title: 金融擁抱AI的機會與挑戰
- (20220311)台灣人工智慧學校「智慧金融專班」
Title: 金融擁抱AI的機會與挑戰
- (20220310)SAS 媒體見面會
Title: 玉山打造 ModelOps 流程,實現AI規模化應用
- (20220218)台灣人工智慧學校【AI 策略治理與案例分析董總班】
Title: 資料治理與AI落地服務
- (20220117)勤業眾信內部知識分享會
Title: 玉山金控的AI轉型之路
2021
- (20211224、Friday)清華大學 AI Seminars (1.5 hour talk)
主辦單位:清華大學
邀請人:賴尚宏教授
Title: 金融擁抱 AI 的機會與挑戰 - AI如何成就金融?
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速進行中,同時也造成AI的技術被無遠弗屆地應用在金融業。此次演講將說明 AI 在金融業的不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如AML黑名單偵測、票據影像辨識、信用卡盜刷偵測、房屋自動估價、語音辨識與合成、智能客服的關鍵應用等,以及因應數位行為而生之風控、監理等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠在實際應用場景順利落地並持續精進的最重要關鍵。
- (20211222)2021安富論壇《房地產金融數位轉型》
Title: 金融機構的AI應用與數位轉型分享
- (20211221)南山人壽 2022 Open Pathways 數位賦能講座
Title: 玉山銀行的轉型經驗—如何以人為本, 用科技引領創新, 看見未來
- (20211214)2021 New Futures期貨學術與實務交流研討會
Title: FinTech新金融服務及AI風控運用與創新
- (20211023、Saturday)國立中山大學【後EMBA】西灣生生塾 (6-hour class)
主辦單位:中山大學
邀請人:中山大學管理學院 黃三益院長
Title: AI 的簡介以及在金融科技與多媒體的應用
Abstract: 本課程將先簡介 AI 的原理及其應用,包含目前的難題與未來的趨勢,最後落地到 AI 在金融產業所扮演的角色及議題(如數位轉型、資料治理、人才培育等),以及在多媒體方面(多用於製造業)的各項前瞻應用。
- (20211016、Friday)台大CASE探索講座 (120-min talk)
主辦單位:台大
邀請人:台大科學教育發展中心(Center for the Advancement of Science Education, CASE)
Title: AI 如何解構、轉化與再造音樂
Abstract: 在AI技術的推波住瀾下,音樂已經走入一個截然不同的世界,此演講將說明如何以AI來解構音樂(人聲分離、鼓聲分離等)、轉化音樂(和弦辨識、哼唱轉譜、歌詞對位等),以及再造音樂(樂器學習、歌聲合成等),並展示如何結合這些技術來產生各項教育與娛樂的應用,例如卡拉OK等。
Video: https://www.youtube.com/watch?v=SUaBUS20N7w
- (20211013、Wed)Visionary talk at ITC (30-min talk)
主辦單位:ITC (International Test Conference )
Title: Machine Learning and Corpus Design for EDA and Beyond
Abstract:
We have been building faster computers to host innovative ML (machine learning) applications such as image recognition and language understanding. But can we reverse the roles and use ML to help EDA (electronic design automation) in order to create better performing chips and computers? The answer is definitely yes. In this talk, we shall cover two such examples, including wafer failure analysis and side channel attacks using ML. Actually these two examples only scratch the surface since there are a number of ML tasks for EDA to produce better computers for advanced ML applications. We believe such positive feedback in the loop will advance both ML technologies as well as chip/computer performance. Moreover, we shall also address the guidelines for ML corpus design, and touch on corpus-hungry ML applicatons that outperform humans by a large margin and were unthinkable before the current AI era.
Video: https://www.youtube.com/watch?v=_1v_r3vKL3o
- (20211008、Fri)2021 台灣人工智慧年會暨 AI 技術應用論壇專題演講 (25-min talk)
主辦單位:台灣人工智慧學會
Title: AI專案在金融場景的美麗與哀愁(The beauty and sorrow of AI projects in financial scenes)
Abstract:
AI在金融場景的應用面可以說是無遠弗屆,其重要性也已經是眾所皆知,但是實際在執行AI專案時,會碰到什麼問題呢?我們將以幾個AI專案為例,來說明這些專案從構想、設計、建模,到落地、維運之間所經歷的各種美麗(願景)與哀愁(痛點)。這些不會在教科書被提到的問題,才是決定某一項AI專案是否能夠在應用場景順利落地並持續精進的關鍵。
The application of AI in financial scenarios can be said to be far-reaching, and its importance is well known, but what problems will actually be encountered when implementing AI projects? We will take several AI projects as examples to illustrate the beauty (vision) and sorrow (pain points) experienced by these projects from conception, design, modeling, to implementation, and maintenance. These issues that will not be mentioned in the textbook are the key to determining whether an AI project can be successfully implemented and continuously improved in application scenarios.
- (20210924、Fri)Ansys 專題演講 (60-min talk)
主辦單位:Ansys
邀請人:Norman Chang
Title: Machine Learning for Audio Processing - Practice and Examples
Abstract:
This talk will introduce machine-learning-based audio enabling technologies that can be used in numerous applications, ranging from speech/music to audio processing in general. One of the technologies is sound source separation, which is a difficult yet important basic skill that can be applied in almost all audio processing scenarios, including speech enhancement and vocal separation from music. Once the sound sources are separated, all subsequent tasks can be greatly simplified, such as audio-based anomaly detection (for engine noise diagnosis, and speaker inspection, etc.), vocal-lyrics alignment, speaker diarization and recognition, and many more. Due to the rapid progress of machine learning, audio tasks that were previously considered impossible are now realized one by one.
本演講將介紹數個基於機器學習的音訊技術,這些技術可以用在多種不同的應用上面、包含語音和音樂的處理。以音源分離為例,這是一項困難但是卻很重要的基本功夫,可以被用在語音辨識內的訊號增強,以及音樂處理內的人聲分離。一旦進行音源分離後,後續所有的分析工作就可以大幅簡化。其他相關的應用還包含以音訊為主的異例偵測、和旋辨識、歌聲和歌詞對位、語者分段及辨識、消除殘響等各種應用。由於機器學習的快速進展,以前被認為不可能的任務,現在都逐一實現了。
- (20210818、Wed)科技部「數位經濟技術創新研發與應用」成果展專題演講 (60-min talk)
主辦單位:科技部
邀請人:陳銘憲副校長
Title: 金融擁抱 AI 的機會與挑戰 - AI如何成就金融?
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速進行中,同時也造成AI的技術被無遠弗屆地應用在金融業。此次演講將說明 AI 在金融業的不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如AML黑名單偵測、票據影像辨識、信用卡盜刷偵測、房屋自動估價、語音辨識與合成、智能客服的關鍵應用等,以及因應數位行為而生之風控、監理等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠在實際應用場景順利落地並持續精進的最重要關鍵。
Video: https://www.youtube.com/watch?v=I3L5i2yh9Jw
- (20210816、Mon)台北市進出口商業工會專題演講 (60-min talk)
主辦單位:台北市進出口商業工會
Title: 疫後新經濟-如何以人工智慧提升企業競爭力
Abstract:
1.人工智慧的機會與挑戰
2.人工智慧產業案例
(1)電腦視覺:無人商店、生態調查、藥品盤點、災防辨識等
(2)語音辨識:遠端醫療、聊天機器人、零接觸金融等
(3)生物認證:資訊安全、智慧門禁等
(4)搜尋與推薦:電商的顧客行為分析等
(5)時間序列分析:商品價格預測、大宗原物料價格預測等
3. 個人和企業如何開始佈署人工智慧
- (20210618、Friday)科技部2021春季「展望」系列演講講座 (120-min talk)
主辦單位:台大天文物理研究所
邀請人:臺灣大學物理學系 張顏暉教授
Title: 金融擁抱 AI 的機會與挑戰 - AI如何成就金融?
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速進行中,同時也造成AI的技術被無遠弗屆地應用在金融業。此次演講將說明 AI 在金融業的不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如AML黑名單偵測、票據影像辨識、信用卡盜刷偵測、房屋自動估價、語音辨識與合成、智能客服的關鍵應用等,以及因應數位行為而生之風控、監理等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠在實際應用場景順利落地並持續精進的最重要關鍵。
Video: https://m.youtube.com/watch?v=qlyXXfsiOBs
- (20210529、Saturday)交通大學 全球領袖班 (120-min talk)
主辦單位:交通大學 全球領袖班
邀請人:交通大學財資系主任 戴天時教授
Title: 金融擁抱 AI 的機會與挑戰
- (20210509、Sunday)政治大學 EMBA 金融科技趨勢與創新課程 (80-min talk)
主辦單位:政治大學 EMBA
邀請人:政大金融科技研究中心主任 王儷玲教授
Title: AI 與金融科技之發展與應用
Abstract:
- (20210424、Saturday)北科大「智慧鐵道系統」課程 (3-hour` talk)
主辦單位:北科大 智慧鐵道產業人才學院
邀請人:北科大機械系 陸元平教授
Title: 人工智慧與大數據分析
- (20210411、Sunday)2021 科研發光 (30-min talk)
主辦單位:科技部 x 台師大 x 科學人雜誌
邀請人:台師大光電所 楊承山教授
Title: AI用於音樂與語音的處理
Abstract: 本演講將說明AI在音訊的應用,包含在音樂方面的辨識與音源分離,以及語音的降噪與合成,現場將穿插各項展示,讓聽眾能夠實際感受AI帶來的各項突破。
Video: https://www.youtube.com/watch?v=MnK1TdjA7nQ
2020
- (20201218)中國電機工程學會學術研討會 (50-min talk)
Title: 金融擁抱AI的機會與難題
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速中。講者將以金融業為例,說明 AI 在不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如信用卡盜刷偵測、智能客服的關鍵應用,以及因應數位行為而生之風控、偵測等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠發揮其功能的重要關鍵。
- (20201211)人工智慧與金融科技研討會 (60-min talk)
Title: 機器學習於金融產業之落地應用的現實考量
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速中。講者將以金融業為例,說明 AI 在不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如信用卡盜刷偵測、智能客服的關鍵應用,以及因應數位行為而生之風控、偵測等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠發揮其功能的重要關鍵。
- (20201125)清華大學智慧校園講座系列 (80-min talk)
Title: 金融擁抱 AI 的機會與挑戰
Abstract: 受到新冠肺炎疫情影響,企業數位轉型的步伐與零接觸經濟的推動正在持續加速中。講者將以金融業為例,說明 AI 在不同面向所扮演的關鍵角色,以及所面臨之挑戰,並進行實例分享,例如信用卡盜刷偵測、智能客服的關鍵應用,以及因應數位行為而生之風控、偵測等需求。此外,我們也會談到資料的治理、模型的管理與精進等,這些維運面的問題通常不會在教科書被提到,但卻是決定某一項AI技術是否能夠發揮其功能的重要關鍵。
- (20201113)台灣人工智慧年會 (50-min keynote speech)
Title: 科技賦能、智慧領航
Abstract: 受到新冠肺炎疫情刺激下,也加速了企業數位轉型的步伐與零接觸經濟的推動。講者將以金融業為例,分享 AI 在其中所扮演之關鍵角色,像是針對大量紓困案件所做快速應變、智能客服的關鍵應用與發展、以及因應數位行為而生之風控、偵測需求等實例分享。期望能協助企業以智能掌舵、用智慧領航。
- (20200930)台大學務處知識分享 (90-min talk)
Title: 人工智慧的簡介與應用
- (20200910)DataCon.TW 2020 (40-min talk)
Title: Common Caveats of Machine Learning
Abstract: Instead of introducing the newest technologies in ML, this talk will pull us back and review some of the common caveats when we are in the routines of ML practice. Most of these caveats occur when the dataset is small or imbalanced, or when the dimension is large. Moreover, most of the caveats can be detected in advance by exerting more caution in coding. We will give some examples of how these caveats might occurs and how to prevent them . By knowing these caveats, you will be able to have a complete view of ML without falling into the fallacy of it.
- (20200729)AI金融科技協會 (50-min talk)
Title: 人工智慧簡介及其在金融科技的應用
- (20200219)首都扶輪社 (90-min talk)
Title: 人工智慧的應用與展示
- (20200109)國際扶輪 3502 地區 人工智慧論壇 (50-min talk)
Title: 人工智慧在多媒體的應用
2019
- (20191229)國泰 Maker in CathayTech 數位科技知識分享會 (40-min talk)
Title: 聊天機器人簡介
- (20191228)台大田徑隊70週年大會的演講 (60-min talk)
Title: 人工智慧的介紹、應用與展示
- (20191220)「商週 CEO 學院:數位領導力」演講 (120-min talk)
Title: 人工智慧在企業的應用、AI實戰案例分享
- (20191018)此為應邀於華碩科技的演講 (60-min talk)
Title: 機器學習用於音樂處理
Abstract: 本演講將說明如何應用機器學習於音樂處理的各項工作,包含哼唱選歌、音訊指紋辨識、人聲分離、節拍追蹤等,並說明相關可能的應用及展示。
- (20190928)台灣消化醫學週會議(2019 Taiwan Digestive Disease Week, TDDW)(25-min talk)
Title: Machine Learning for Precision Medicine: Practical Concerns and Case Studies
Abstract: This talk introduce the common practice of machine learning for medical/healthcare data analytics, including practical concerns and difficulty in data cleaning and normalization. We shall use a case of fatty liver detection to illustrate how these techniques can help the creation of a better model for better prediction.
- (20190926)臺大工業智慧製造研討會 (40-min talk)
Title: 機器學習於智慧製造的應用案例分享:晶圓錯誤型態辨識與機台異音分類
- (20190925)此為應邀於「2019 TANET」的 keynote speech (50-min talk)
Title: 機器學習用於音樂處理及其延伸應用
Abstract: 本演講將說明如何應用機器學習於音樂處理的各項工作,包含哼唱選歌、音訊指紋辨識、人聲分離、節拍追蹤等,並說明相關可能的應用及展示。同時我們也說明這些方法如何延伸到其他面向的應用,包含金融時間序列預測及工廠異常檢測等。
- (20190906)創業家兄弟 (50-min talk)
Title: 人工智慧的現況與挑戰
- (20190904)祥碩科技 (3-hour talk)
Title: 人工智慧導論
- (20190801)「商周 CEO 學院:數位領導力中華郵政專班」演講 (120-min talk)
Title: 人工智慧概論、人工智慧於金融科技之應用
- (20190704)此為應邀於「2019 MATLAB人工智慧與金融科技論壇」的演講 (40-min talk)
Title: 人臉聲紋在智能客服的應用
Abstract: 本演講說明人臉辨識與聲紋比對在金融科技的應用,例如智能客服及精準行銷,並分享我們實驗室在這方面的研究成果。
- (20190703)此為應邀於北一女的演講 (40-min talk)
Title: 洗耳恭聽的AI
Abstract: 如何讓電腦可以聽得懂人類所說的話,是目前最夯的應用,例如手機上面的語音助理,智慧家庭中的智慧音箱應用等等。在今天的分享裡面,我們將為各位同學介紹人工智慧是利用什麼方式去解析語音以及各種音訊的資料,例如:如何聽得懂語音命令、如何從流行音樂抽出清唱歌聲等,這些應用都讓電腦成為我們更厲害的一雙耳朵。
- (20190516 開始)此為應邀於世界先進的短期課程,為期四週,每週三個小時。
Title: 機器學習簡介、機器學習用於智慧製造的案例分享
2018
- (20181122)此為應邀於「2018 健康及生醫大數據研討會」的專題演講 (20-min talk)
Title: 機器學習應用案例分享
Abstract: 本演講將說明在科技部贊助的台大醫神計畫之下,所開發出來的幾項機器學習應用案例,包含CPR病人偵測、胸部X光病歷結構化,以及呼吸週期辨識。本演講將說明在開發這幾項應用時所遇到的困難及解決方式,以及最後的效能評估。
- (20181020)此為應邀於「2018 MATLAB/Simulink Tech Forum」的專題演講 (45-min talk)
Title: 中文斷詞的原理及應用, Basics and Applications of Chinese Word Segmentation
Abstract: 本演講將介紹中文斷詞的 基本原理及方法,並展示MATLAB的實現及效能比較,最後並以實例說明中文斷詞如何應用在文件自動分類、意見探勘、中文文字轉語音、語音辨識等各種重要的領域。
- (20181004)此為應邀於「2018 Ansys Convergence Conference at Taiwan」的專題演講 (30-min talk)
Title: Case Studies for Machine Learning Used in Intelligent Manufacturing and EDA
Abstract: 本演講將分享機器學習用於半導體製造業及EDA的各種實際應用。
- (20180818)人工智慧學校, 中研院 (45-min talk)
Title: 機器學習用於製造與行銷的案例分享
Abstract: 本演講將分享幾個機器學習的案例,包含晶圓錯誤樣式辨識、使用SEM影像的深度重建,以及用於精準行銷的顧客行為分析。我們將介紹如何使用機器學習於這些應用,並說明如何進行特徵抽取、模型選取以及效能分析等,以優化這些應用並朝向智慧型製造與行銷的目標邁進。
- (20180712)Talk for Ansys (60-min talk)
Title: Ensemble Learning
Abstract: This talks cover the basics of ensemble learning, including random forests and extreme gradieng boosting, including XGBOOST which has won several data science competitions, such as Kaggle.
- (20180117)ANSYS, Inc. 2645 Zanker Road San Jose CA 95134 (1.5-hour talk)
Title: A Canonical Approach to Machine Learning and Its applications
Abstract: In this talk, I will review the basic steps in machine learning, and explain how to carry out a canonical approach to this end. I will use a small yet representative task to demonstrate how the canonical approach can be use for any specific applications. Moreover, I will also cover several on-going projects being conducted in our lab to demonstrate example industrial applications of machine learning.
Bio: Roger Jang received his Ph.D. from EECS Department at UC Berkeley. He studied fuzzy logic and artificial neural networks with Prof. Lotfi Zadeh, the father of fuzzy logic. As of Sept. 2017, Google Scholar shows more than 13,000 citations for his seminal paper on ANFIS (adaptive neuro-fuzzy inference systems), which is a popular paradigm combining neural networks and fuzzy logic for regression tasks in machine learning. After obtaining his Ph.D., he joined the MathWorks to coauthor the Fuzzy Logic Toolbox (for MATLAB). He has since cultivated a keen interest in implementing industrial software for pattern recognition and computational intelligence. He has published one book entitled "Neuro-Fuzzy and Soft Computing" by Prentice Hall. He has also maintained toolboxes for "Machine Learning" and "Speech and Audio Processing". He is the general chair of ISMIR (International Society for Music Information Retrieval) Conference, Taipei, 2014 and a general co-chair of ISMIR Conference, Suzhou, 2017. His research interests include machine learning and data analytics, with applications to speech scoring, music retrieval, fintech, and semiconductor manufacturing intelligence. He is currently a professor of CSIE Dept. at National Taiwan University. He is also the director of the IT Office at National Taiwan University Hospital. For further information on Prof. Jang, please visit "http://mirlab.org/jang".
2017
- (20171109)人工智慧年會專題演講 (45-min talk)
Title: 音樂檢索與歌聲抽取
Abstract: 本演講將回顧音樂檢索的過去與現況,特別是在哼唱選歌及音訊指紋辨識這兩個領域,並說明目前音樂檢索所碰到的最大挑戰。針對這個挑戰,我們將解釋如何從複音音樂進行歌聲抽取以及其重要性。在不同的應用情境下,我們使用的方法包含深度神經網路以及主動是噪音消除,同時我們也將說明如何將抽取出來的歌聲用於各項音樂相關的應用,包含哼唱選歌、口水歌辨識、歌詞對位、歌聲評分等,現場並會進行各項相關展示。
Bio: 張智星教授於 1992 年取得加州大學柏克萊分校的電機電腦博士,博士論文即在探討模糊邏輯與類神經網路的建模與迴歸,1993 年的單一作者論文「ANFIS: Adaptive-network-based fuzzy inference system」開啟模糊推論系統自我學習的大門,Google Scholar 被引用數達到 12740(2017/07/31)。張教授於 1993 年加入美國 MathWorks 公司,開發模糊邏輯工具箱。1995 年回台任教後,研究方向轉向機器學習的各項應用,包含語音評分、音樂檢索、文件分類、影像辨識等領域。他曾經擔任 2014 年 ISMIR(在台北舉行)的 General Chair 以及 2017 年 ISMIR(在蘇州舉行)的 General Co-chair,他的團隊也在歷年 MIREX 國際音樂檢索評比中的數個項目拿下第一名的佳績。張教授著重理論和實作的整合,曾經將所開發的語音評分、音樂檢索的程式庫授權給多家知名廠商,同時也執行多項產學合作計畫,並擔任工研院、資策會及其它相關廠商顧問,產學合作績效亮麗。
- (20171021)The fifth China Conference on Sound and Music Technology (CSMT 2017)(1-hour talk)
Title: Music Information Retrieval and Singing Voice Separation
Abstract: This talk will review the current status of music information retrieval (MIR), especially the most successful two paradigms of MIR, query by singing/humming and audio fingerprinting. We will also address the most challenging talks in MIR, which leads to the task of singing voice separation from monaural audio music, using the methods of DNN (deep neural networks) and ANC (active noise cancellation) under various application scenarios. We will explain how singing voice separation can be applied to various aspects of MIR, including query by singing/humming, cover song identification, lyrics alignment, and singing scoring.
- (20170926)ANSYS, Inc. 2645 Zanker Road San Jose CA 95134 (1-hour talk)
Title: Machine Learning for Manufacturing Intelligence – Case studies
Abstract: In this talk, I will start with a brief introduction to our lab at National Taiwan University, including our research directions and current projects. Then I’ll focus on case studies of using machine learning for semiconductor manufacturing intelligence, including wafer failure map classification, depth reconstruction via SEM images, and predictive maintenance. I will explain how to apply machine learning techniques to such tasks, including feature extraction, model selection, input selection/extraction, and performance evaluation.
Bio: Roger Jang received his Ph.D. from the EECS Department at UC Berkeley. He studied fuzzy logic and artificial neural networks with Prof. Lotfi Zadeh, the father of fuzzy logic. As of Sept. 2017, Google Scholar shows around 13,000 citations for his seminal paper on adaptive neuro-fuzzy inference systems (ANFIS). After obtaining his Ph.D., he joined the MathWorks to coauthor the Fuzzy Logic Toolbox (for MATLAB). He has since cultivated a keen interest in implementing industrial software for pattern recognition and computational intelligence. He is currently a professor in the CSIE Dept. of National Taiwan Univ., Taiwan. He has published one book entitled "Neuro-Fuzzy and Soft Computing" by Prentice Hall. He has also maintained toolboxes for "Machine Learning" and "Speech and Audio Processing". He is the general chair of ISMIR (International Society for Music Information Retrieval) Conference, Taipei, 2014 and will be a general co-chair of ISMIR Conference, Suzhou, 2017. His research interests include machine learning and data analytics, with applications to speech scoring, music retrieval, fintech, and semiconductor manufacturing intelligence. For further information on Prof. Jang, visit "http://mirlab.org/jang".
- (20170803)墾丁福華飯店舉行的 2017 超大型機體電路設計暨計算機輔助設計技術研討會的45分鐘演講。
Title: 機器學習案例分享
Abstract: 本演講將分享幾個用於半導體製造的機器學習的案例,包含晶圓錯誤樣式辨識、使用SEM影像的深度重建,以及自動駕駛設計。我們將介紹如何使用機器學習於這些應用,並說明如何進行特徵抽取、模型選取以及效能分析等,以優化這些應用並朝向智慧型製造的目標邁進。
2016
- (20161209)此為應邀於清華大學電機系的專題演講。
Title: Music Information Retrieval (MIR) – Overview and Challenges
Abstract: This talk will give a brief overview of MIR, and explain the basics of QBSH (query by singing/humming) and AFP (audio fingerprinting) with live demos. Moreover, we shall address our recent progress on singing voice separation, and point out the current challenges and future directions of MIR.
- (20161026)此為應邀於 MATLAB/Simulink Tech Forum 的專題演講。
Title: Speech and Singing Voice Enhancement via DNN
Abstract: This talk will presents several approaches of using DNN for speech and singing voice enhancement. In particular, we shall address the problem of speech/singing voice enhancement via two scenarios, including active noise cancellation, and extraction from monaural recordings. Demos will be given to show the results of DNN for such tasks.
- (20160919)此為應邀於 Tohoku University, Sendai, Japan 舉行的 Association of East Asian Research Universities (AEARU): Web Technology and Computer Science Workshop 2016 (WTCS2016) - Computer Science and Data Science 的半小時邀請演講。
Title: Overview of Music Information Retrieval
Abstract: This talk presents an overview of music information retrieval (MIR), with an emphasis on the use of deep neural networks (DNN) to achieve state-of-the-art performance in singing voice separation.
- (20160909)此為應邀於新竹國賓飯店舉行的 e-Manufacturing & Design Collaboration Symposium 的 1-hour tutorial。
Title: Case Studies of Machine Learning for Manufacturing Intelligence
Abstract: This talk focuses on case studies of using machine learning for semiconductor manufacturing intelligence. In particular, we shall introduce three important tasks in semiconductor manufacturing, including wafer map classification, depth reconstruction via SEM images, and predictive maintenance. We shall explain how to apply machine learning to such tasks, including feature extraction, model selection, input selection/extraction, and performance evaluation.
- (20160728)此為應邀於台大醫院資訊室的一小時演講。
Title: 音樂資訊檢索簡介 (Introduction to Music Information Retrieval)
Abstract: 本演講將介紹音樂資訊檢索的現況與未來,包含此領域用到的技術以及碰到的難題。演講中將穿插各項展示,例如哼唱選歌、音訊指紋辨識、節拍辨識、曲風分類、即時音高調整、歌聲抽取等,以讓聽眾瞭解音樂分析及檢索的實際應用面。
- (20160422)此為應邀於「ISAC 大專校院資訊人員研討會」演講。
Title: 機器學習及多媒體資訊檢索用於大資料分析的產學合作案例分享
2014
- (20141219)"Music Information Retrieval - Overview and Challenges":
此為應邀於 2014/12/19 於中研院舉行的 IR Workshop 的 keynote speech。
- (20141023)MATLAB/Simulink Tech Forum:機器學習用於半導體製造智慧 (Machine Learning for Semiconductor Manufacturing Intelligence)
摘要:本演講將簡介機器學習工具箱的功能與特性,並說明此工具箱在半導體晶圓製造的實際應用,包含晶圓缺陷型態分類、機台維修階段預測等,以使聽眾瞭解機器學習在產業界的實際應用面。
Abstract: This talk will give a brief introduction to the Machine Learning Toolbox, and describe several intelligent applications in semiconductor manufacturing, including wafer failure pattern classification and machine maintenance stage prediction, such that the audience can acquire a full understanding of the importance of machine learning in semiconductor manufacturing intelligence.
- (20140902)Kyoto University: Music Information Retrieval - Overview and Challenges
Abstract: This talk will introduce music information retrieval in general, and explain the basics of QBSH (query by singing/humming) and AFP (audio fingerprinting) in particular. Moreover, we shall give various demos of QBSH and AFP in practice, and point out the current challenges and future directions of music information retrieval.
- (20140604)Academia Sinica: Recent Improvement Over QBSH and AFP
This talk introduces the basics of QBSH (query by singing/humming) and AFP (audio fingerprinting), and describe our recent improvements over these two classical ways of music information retrieval.
2013
- (20131220)台南大學:機器學習的實際應用 (Machine Learning in Action)
摘要:本演講將介紹機器學習工具箱的功能與特性,並說明其相關應用,特別是在半導體晶圓廠的實際應用,包含晶圓缺陷型態分類、預測機台維修階段等,以使聽眾瞭解機器學習在產業界的應用面。
- (20131023)MATLAB/Simulink Tech Forum:機器學習的實際應用 (Machine Learning in Action)
摘要:本演講將介紹機器學習工具箱的功能與特性,並說明其相關應用,特別是在半導體晶圓廠的實際應用,包含晶圓缺陷型態分類、預測機台維修階段等,以使聽眾瞭解機器學習在產業界的應用面。
- (20130916-0920)A short course on "QBSH and AFP as Two Successful Paradigms of Music Information Retrieval" at RuSSIR 2013, Kazan, Russia,
- (20130524)逢甲大學 MATLAB DAY:機器學習的實際應用 (Machine Learning in Action)
摘要:本演講將介紹機器學習工具箱的功能與特性,並說明其相關應用,特別是在半導體晶圓廠的實際應用,包含晶圓缺陷型態分類、預測機台維修階段等,以使聽眾瞭解機器學習在產業界的應用面。
2012
- (20120102)台大電機所:音訊辨識技術與應用 - 以遊戲為導向的語音與音樂學習
摘要:本演講將說明各項音訊辨識技術在語音與音樂學習方面的應用,這些技術包含語音識別(speech recognition)、文字轉語音(text-to-speech conversion)、語音評分(speech assessment)、哼唱選歌(query by singing/humming)、音高追蹤(pitch tracking)、節拍追蹤(beat tracking)、曲風分類(music genre classification)等。由於每個技術項目都有各自的特性,因此我們在應用於以遊戲為導向的學習時,必須考慮到這些特性,才能建構出有趣的應用程式。本演講將穿插各項展示,讓聽眾能夠體驗每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
- (20120229)北京清華大學訊息所:音訊辨識技術與應用 - 以遊戲為導向的語音與音樂學習
- (20120405)東南大學電機系:影像辨識與應用
- (20120406)HMI創新應用與技術研討會:機器學習與音訊辨識在HMI的整合應用
- (20120412)東南大學電機系:音訊辨識與應用
- (20120503)2012 Android Days技術交流大會:機器學習與音訊辨識在HMI的整合應用
- (20120618)2012 GPU Workship Taiwan:Query by Singing/Humming: A Practical Application of GPU
- (20120630)語音訊號處理研討會:Two Paradigms for Music IR: Query by Singing/Humming and Audio Fingerprinting
- (20121025)MATLAB/Simulink Tech Forum:機器學習的實際應用 (Machine Learning in Action)
摘要:本演講將介紹機器學習工具箱的功能與特性,並說明其相關應用,特別是在半導體晶圓廠的實際應用,包含晶圓缺陷型態分類、預測機台維修階段等,以使聽眾瞭解機器學習在產業界的應用面。
- (20121226)元智大學化材系:機器學習的實際應用 (Machine Learning in Action)
摘要:本演講將介紹機器學習工具箱的功能與特性,並說明其相關應用,例如人臉辨識、哼唱選歌等。此外,我們會特別說明在半導體晶圓廠的實際應用,包含晶圓缺陷型態分類、預測機台維修階段等,以使聽眾瞭解機器學習在現今產業界可能發揮的面向和相關的效能。
2011
- (20111228)中央研究院資訊科學研究所,音樂資訊檢索暨社群服務技術研討會(Workshop on Music Information Retrieval and Social Network Service):Query by Singing/Humming: An Overview
摘要:This talk gives an overview of QBSH (query by singing/humming), including its history, methodologies, commercial systems/demos, and trends/challenges.
- (20110610)台大資管系:音訊辨識技術與應用 - 以遊戲為導向的語音與音樂學習
摘要:本演講將說明各項音訊辨識技術在語音與音樂學習方面的應用,這些技術包含語音識別(speech recognition)、文字轉語音(text-to-speech conversion)、語音評分(speech assessment)、哼唱選歌(query by singing/humming)、音高追蹤(pitch tracking)、節拍追蹤(beat tracking)、歌聲分離(singing voice separation)等。由於每個技術項目都有各自的特性,因此我們在應用於以遊戲為導向的學習時,必須考慮到這些特性,才能建構出有趣的應用程式。本演講將穿插各項展示,讓聽眾能夠體驗每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
- (20110105)交大資工系:音訊辨識技術與應用 - 以遊戲為導向的語音與音樂學習
摘要:本演講將說明各項音訊辨識技術在語音與音樂學習方面的應用,這些技術包含語音識別(speech recognition)、文字轉語音(text-to-speech conversion)、語音評分(speech assessment)、哼唱選歌(query by singing/humming)、音高追蹤(pitch tracking)、節拍追蹤(beat tracking)、歌聲分離(singing voice separation)等。由於每個技術項目都有各自的特性,因此我們在應用於以遊戲為導向的學習時,必須考慮到這些特性,才能建構出有趣的應用程式。本演講將穿插各項展示,讓聽眾能夠體驗每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
2010
- (20101227)資策會:音訊辨識技術與應用 - 以遊戲為導向的語音與音樂學習
摘要:本演講將說明各項音訊辨識技術在語音與音樂學習方面的應用,這些技術包含語音識別(speech recognition)、文字轉語音(text-to-speech conversion)、語音評分(speech assessment)、哼唱選歌(query by singing/humming)、音高追蹤(pitch tracking)、節拍追蹤(beat tracking)、歌聲分離(singing voice separation)等。由於每個技術項目都有各自的特性,因此我們在應用於以遊戲為導向的學習時,必須考慮到這些特性,才能建構出有趣的應用程式。本演講將穿插各項展示,讓聽眾能夠體驗每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
- (20101026)MATLAB Technical Forum: Speech and Audio Processing with MATLAB
Abstract: In this talk, I will briefly review several machine learning techniques for data clustering and pattern recognition. In particular, I will cover K-means clustering, hierarchical clustering, KNNC (k-nearest neighbor classifier), quadratic classifier, and GMM (Gaussian mixture model) classifier. We shall focus on how to implement such tasks, and how to effectively visualize the clustering/classification results with MATLAB.
- (20100507)2010 民生電子論壇─多媒體與數位生活的饗宴:音樂資訊檢索的現況與未來
- (20100504)2010 語音訊號處理研討會(Advanced Speech Enabling Human-Machine Interface):語音評分的方法與應用
2009
- (20091020)MATLAB Technical Forum: Machine Learning for Data Clustering and Pattern Recognition
Abstract: Speech and Audio have been an important part of multimedia processing. In this talk, we shall cover important aspects of speech/audio processing and recognition using MATLAB. In particular, we shall address the applications of speech/speaker recognition, speech assessment, and query by singing/humming, and highlight how MATLAB/Simulink can be used effectively for data analysis/visualization for such applications.
2008
- (20080613)資策會:數位學習領域之語音技術應用分享(語音辨識,電腦輔助練習,互動式電子寵物)
- (20080714)北京大學計算語言所:語音技術在口說華語學習的應用
摘要:本演講將說明各項語音技術在華語學習方面的應用,這些語音技術包含語音識別(speech
recognition)、文字轉語音(text-to-speech conversion)、語音評分(speech
assessment)、聲調辨識(tone recognition)、韻律轉換(prosody conversion)等,由於每個技術項目都有各自的特性,因此我們在應用於口說華語學習時,必須考慮到這些特性,才能相輔相成。本演講將穿插各項演示,讓各位老師同學能夠體驗每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
- (20080715)北京微軟亞洲研究院:Progressive Filtering and Its Application for Query-by-Singing/Humming
Abstract: This talk presents the mathematical formulation and design methodology of progressive filtering (PF) for multimedia information retrieval, and reports its application on the so-called query by singing/humming (QBSH), or more formally, melody recognition. The concept of the proposed PF and the corresponding efficient design method based on dynamic programming are applicable to large multimedia retrieval systems for striking a balance between efficiency (in terms of response time) and effectiveness (in terms of recognition rate). The application of the proposed PF to a 5-stage QBSH system is reported, and the experimental results demonstrate the feasibility of the proposed approach.
- (20080729)北京微軟亞洲研究院:Research Activities at MIR Lab
- (20080731)北京三星研究院:哼唱選歌與語音評測
- (20081121)資策會:語音技術在口說華語學習的應用
內容摘要:本演講將說明各項語音技術在華語學習方面的應用,這些語音技術包含語音辨識(speech recognition)、文字轉語音(text-to-speech conversion)、語音評分
(speech assessment)、聲調辨識(tone recognition)、韻律轉換(prosody
conversion)等,由於每個技術項目都有各自的特性,因此我們在應用於口說華語學習時,必須考慮到這些特性,才能相輔相成。本演講將穿插各項展示以便說明每項技術的優點和缺點,並說明如何以流程和創意來發揮最大的學習效果。
2005
- (20051105)台中一中:音訊技術在娛樂、教育與家電的應用
摘要:
隨著電腦運算速度的突飛猛進,許多需要大量資料處理與運算的音訊技術,漸漸滲入各個領域,原先高不可攀的理論,變成隨處可見的應用。本演講將介紹基本的音訊處理,並說明這些技術在娛樂、教育與家電的應用,演講中會穿插各種展示,以讓同學們瞭解音訊技術的實際應用面,包含:
- 互動式電腦/手機卡拉OK:多模式音樂檢索、歌聲即時評分
- 電腦輔助口說英語/漢語學習:語音辨識、音高追蹤
- 背書機:語音辨識與合成
- 科南機:語音轉換與變調
- 大頭狗與頑皮豹:語音辨識、旋律辨識
2004
- (20041210)電機工程學會年會:多模式音樂檢索在數位生活的應用(隨著電腦速度的加快與多媒體處力能力的提高,在電腦上進行卡拉OK歡唱已經不是遙不可及的夢想,尤其再加上各種音訊處理與辨識技術,以及社群與互動模式的建立,卡拉OK會漸漸變成一項互動式的線上遊戲產業,扮演個人電腦與數位家庭在娛樂面的重要角色。本演講將說明如何應用多模式音樂檢索於電腦卡拉OK,以及此種應用模式對於數位生活所帶來的衝擊與影響。(在演講中會穿插系統展示,包含卡拉OK軟體與智慧型互動玩具。)
本演講的大綱如下:
- 互動式電腦卡拉OK的簡介
- 技術面:語音辨識、旋律辨識、歌聲即時評分、哼唱譜和弦
- 應用面:玩具、手機、PC、線上遊戲
- 系統展示與未來展望
- (20041030-31)2004清華「知識嘉年華」:互動式電腦卡拉OK:你今天唱了嗎?(隨著電腦速度的加快與多媒體處力能力的提高,在電腦上進行卡拉OK歡唱已經不是遙不可及的夢想,尤其再加上各種音訊處理與辨識技術,以及社群與互動模式的建立,卡拉OK會漸漸變成一項互動式的線上遊戲產業,扮演個人電腦與數位家庭在娛樂面的重要角色。)
- (20040517)台灣科技大學資訊工程系:多模式音樂檢索
- (20040317)交通大學電信工程系:多模式音樂檢索
2003
- (20031113)健康管理學院資訊傳播系:多模式音樂檢索
- (20030527)長庚大學電機工程系:多模式音樂檢索
- (20030521)師範大學資訊工程系:多模式音樂檢索
- (20030324)台灣科技大學電機工程系:多模式音樂檢索
- (20030318)The 4th Sino Franco Workshop on Web Technologies at 淡江大學:An Internet Music Search Enginewith Multi-modal User Interface
2002
- (20020509)交通大學資訊工程系:多模式音樂檢索(本演講將介紹多模式音樂檢索的概念、研究與實作,以及其他與音樂及音訊的相關研究,如歌聲合成、歌聲評分、自動配和弦、語音辨識等。在演講中會穿插系統展示,以使同學瞭解相關技術的實用性。)
- (20020315)清華大學電機工程系:多模式音樂檢索及語音辨識(本演講將介紹多模式音樂檢索的概念、研究與實作,以及其他與音樂及音訊的相關研究,如語音辨識、歌聲合成與評分等。在演講中會穿插系統展示,以使同學瞭解相關技術的實用性。)
2001
- (20011219)中央大學資訊工程系:多模式音樂檢索、歌聲合成與評分(本演講將介紹多模式音樂檢索的概念、研究與實作,以及其他與音樂的相關研究,如歌聲合成與評分等。本演講亦會說明在將研究成果轉化成商品的過程中,所遇到的各種困難以及可能的解決方案。)
- (20011123)成功大學資訊工程系:音樂檢索、歌聲合成與評分
- (20010926)清華大學資訊系碩士班書報討論:音樂檢索與歌聲評分
- (20010804)清華大學科管所 EMBA 專修班:音樂檢索與歌聲合成 - 技術簡介與商業化考量
- (20010713)人工智慧論壇(人工智慧學會主辦)專題演講:音樂資訊檢索
- (20010705)清華大學科學研習營:歌聲辨識、合成與評分
- (20010413)元智大學電機系:「哼唱選歌」系統的研究與實作
- (20010301)交通大學資工系:找歌?用唱的!
2000
- 2000 MATLAB Conference at Taipei: Audio/Melody Recognition Using MATLAB/Simulink
- (20001214)中央研究院資訊所:「以歌選歌」系統的研究與實作(Research/Development of a Content-Based Music Retrieval System)
- 中山科學院:Scientific Visualization with MATLAB
- 交通大學資科系
- 清華大學資訊系碩士班書報討論:Content-based Music Retrieval System (「以歌選歌」系統)
1999
- 1999 MATLAB Conference at Singapore: Content-based Music Retrieval and Speaker Recognition
- 1999 MATLAB Conference at Taipei: Query by Singing
- 清華大學資訊系碩士班書報討論:Pattern Recognition