第21屆國際語音通訊協會年會INTERSPEECH 2020將於2020年10月25日- 30日全程在線舉行。本次大會共邀請到4位重磅嘉賓做大會主題報告(Keynote Talk),涉及的演講內容包括語言建模、語音聽覺感知、口語技術和面向會話應用的底層語音交互技術等四個方面。今天給大家介紹的演講嘉賓是:IEEE Fellow、ISCA Fellow,國立臺灣大學教授Lin-Shan Lee。他將給大家帶來題為「Doing Something we Never couldwith Spoken Language Technologies-from early days to the era of deep learning」的大會報告。
Lin-Shan Lee: 於1974年獲得國立臺灣大學電子工程學士學位、於1977年獲得史丹福大學電子工程博士學位。1982年起擔任國立臺灣大學電子工程與計算機科學教授,曾任國立臺灣大學計算機科學系系主任(1982-1987年)、電子工程與計算機科學學院院長(2009-2012年),臺灣中央研究院信息科學研究所研究員,並擔任研究所所長(1991-1997年)。
Lin-Shan Lee教授研究興趣涵蓋通信、語音語言處理和計算機輔助語言學習多個領域,包括數字傳輸理論、通信信號處理以及語音識別與合成、韻律建模、口語對話、口語內容檢索與理解等研究方向。他開發並發布了最早的並且非常完整的漢語口語處理技術,包括語音合成、自然語言句法和語法分析器以及大詞彙連續語音識別。他還發布了許多世界上最早的漢語口語處理系統,極大地推動了漢語口語處理的發展,包括語音合成系統、自然語言處理分析器,大詞彙量語音識別系統、口語內容檢索系統和口語對話系統。近年來,他在口語內容檢索和瀏覽方面的主要貢獻也得到了全球的廣泛認可。
Lin-Shan Lee教授曾在IEEE通信協會擔任多個職務、包括亞太地區主席(1994-1995)、理事會成員(1995-1997)、國際事務副主席(1996-1997)和獎項委員會主席(1998-1999)。他是2002年在臺北召開的IEEE Globecom的程序委員會主席、他還曾任ISCA董事會成員(2001-2009),IEEE 信號處理協會的傑出講師(2007-2008), IEEE Signal Processing Magazine(2003-2006)和IEEE/ACM Trasactions on Audio,Speech and Language Processing (2012-2013)副主編以及ICASSP2009主席。他在國際頂級會議和期刊上發表了大量的文章,並擁有大量的國內外專利,1992年當選IEEE Fellow,2010年當選ISCA Fellow。
北京時間10月28日18:00到19:00,Lin-Shan Lee教授將發表題為「Doing Something weNever could with Spoken Language Technologies-from early days to the era ofdeep learning」的主題演講,歡迎大家在線觀看。
Title: Doing Somethingwe Never could with Spoken Language Technologies - from early days to the eraof deep learning
Time: Wednesday, 28October, 18:00-19:00 (GMT+8)
Speaker: Lin-shan Lee, National Taiwan University
Abstract:
Some research effort tries to do somethingbetter, while some tries to do something we never could. Good examples for theformer include having aircrafts fly faster, and having images look morebeautiful; while good examples for the latter include developing the Internetto connect everyone over the world, and selecting information out of everythingover the Internet with Google to name a few. The former is always very good,while the latter is usually challenging. This talk is about the latter.
A major problem for the latter is those we couldnever do before was very often very far from realization. This is actuallynormal for most research work, which could be enjoyed by users only after beingrealized by industry when the correct time arrived. The only difference is herewe may need to wait for longer until the right time comes and the rightindustry appears. Also, the right industry eventually appeared at the righttime may use new generations of technologies very different from the earliersolutions found in research.
In this talk I'll present my personalexperiences of doing something we never could with spoken languagetechnologies, from early days to the era of deep learning, including how Iconsidered, what I did and found, and what lessons we can learn today, rangingover various areas of spoken language technologies.
會議時間:2020年10月25日-30日
會議官網:http://www.interspeech2020.org/
更多會議實時信息,請留意會議官方公告。
關注會議網站,請掃描下方二維碼:
關注微信公眾號INTERSPEECH2020,請掃描下方二維碼: