《Speech_Technology》由会员分享，可在线阅读，更多相关《Speech_Technology（44页珍藏版）》请在金锄头文库上搜索。
1、Ubiquitous Computing: A Vision for How Speech Becomes MainstreamKai-Fu Lee Corporate Vice President Microsoft CorporationPresented at SpeechTEK October 29,2002,管理资源吧（），提供海量管理资料免费下载！,Talk Outline,Four trends in the “Digital Decade”. Natural UI (speech & language UI) Vision. Four trends accelerate nat
2、ural UI. A technology roadmap to natural UI.,管理资源吧（），提供海量管理资料免费下载！,Computing Revolutions,管理资源吧（），提供海量管理资料免费下载！,Trends in the Digital Decade: Everything connected,More computing and capacity. Ubiquity of connected devices. Structured content. Distributed computing platform.,管理资源吧（），提供海量管理资料免费下载！,2001
3、 : The Digital Decade,More computing and capacity. Moores Law,CPU,2X / 18 months,Bandwidth,3X / 18 months,Disk Capacity,3X / 18 months,管理资源吧（），提供海量管理资料免费下载！,2001 : The Digital Decade,More computing and capacity Ubiquity of connected devices. PCs, telephones, smart phones, televisions, cars Moores La
4、w applies here! (e.g., PocketPC, TabletPC) Standards on how devices “talk” to each other HTML, HTTP, XML, SOAP, UDDI, WSDL Metcalfes Law: Value of network = nodes2.,管理资源吧（），提供海量管理资料免费下载！,2001 : The Digital Decade,More computing and capacity Ubiquity of connected devices. Structured content. XML = Un
5、iversal standards for describing data. XML makes content “readable” by programs. XML makes content more like databases than text.,管理资源吧（），提供海量管理资料免费下载！,On Pocket PC,On PC,管理资源吧（），提供海量管理资料免费下载！,2001 : The Digital Decade,More computing and capacity. Ubiquity of connected devices. Structured content. D
6、istributed computing Platform. Built on Standards (XML web services).,管理资源吧（），提供海量管理资料免费下载！,Web Software Today,Conventional Browser,Bank Account,管理资源吧（），提供海量管理资料免费下载！,XML Web Services Software,Stock Trading,Personal Finance Portal,Rich Application,XML Web Service,XML Web Service,管理资源吧（），提供海量管理资料免费下载
7、！,2001 : The Digital Decade,More computing and capacity Ubiquity of connected devices. Structured content. Distributed computing platform. Built on standards (XML web services). Transparent to the end-user. One development & execution model. Security and privacy critical.,管理资源吧（），提供海量管理资料免费下载！,Talk
8、Outline,Four trends in the “Digital Decade”. Natural UI (speech & language UI) Vision. Four trends accelerate natural UI. A technology roadmap to natural UI.,管理资源吧（），提供海量管理资料免费下载！,The Vision for “Natural” UI,Users naturally articulate what they mean, on any device, to any application or web service,
9、 and have their intention interpreted and executed accurately. Why NUI? Expressive more powerful. Delegation more efficient. Natural no learning. Scalable any device.,管理资源吧（），提供海量管理资料免费下载！,Natural UI Will Enable,Smart Search Find the Bill Gates book on future Smart Help How do I replace my printer c
10、artridge? Question Answering When is the Britney Spears concert? Commands / Tasks Send flowers to mom on her birthday Pro-Active agent Hold all calls unless its from my family.,管理资源吧（），提供海量管理资料免费下载！,Talk Outline,Four trends in the “Digital Decade”. Natural UI (speech & language UI) Vision. Four tren
11、ds accelerate natural UI. A technology roadmap to natural UI.,管理资源吧（），提供海量管理资料免费下载！,1. Moores Law & Speech,Moores Law helps ASR accuracy Leveraging Moores Law + more data + research. Predictable 10% error reduction or more per year. Human-level performance possible in 10-20 years.,管理资源吧（），提供海量管理资料免费
12、下载！,1. Moores Law & Speech,No need to wait for 10 a 20 years. With real systems, error reduction 10% / year. Most applications dont need human-level performance. Every year, new applications will be enabled: Hierarchical Natural language dialog. Fixed vocabulary Natural dictation. Limited commands “
13、How may I help you.”,“To me, speech recognition will be a transforming capability once it finally comes into being. Im talking about when you can speak to your computer and it will understand what youre saying in context.”Gordon Moore, 2002,管理资源吧（），提供海量管理资料免费下载！,2. Ubiquitous Devices & Speech,管理资源吧（
14、），提供海量管理资料免费下载！,High,Internet TV,Phone,PDA,Ease of text input (keyboard/pen),Ease of GUI (screen/ Pointer),Low,High,PC,Tablet PC,Screen Phone,Screen Phone,PDA,Tablet PC,Car,Car,Internet TV,2. Ubiquitous Devices & Speech,管理资源吧（），提供海量管理资料免费下载！,Opportunities for Speech,Ease of text input (keyboard/pen)
15、,Ease of GUI (screen/ Pointer),High,High,Low,Speech-Only Command/Control,Dictation,Multimodal Command/Control,管理资源吧（），提供海量管理资料免费下载！,2. Ubiquitous Devices & Speech,Increasing number of screen phones. Multimodal opportunity (speech-in, screen-out). Increasing number of keyboard-less devices. Dictation opportunity. Increasing number of mouse-less devices. Multimodal opportunity (speech+pen input).,管理资源吧（），提供海量管理资料免费下载！,Demonstration Speech on TabletPC,
