Speech Recognition: An Historical Perspective on the Business and Technology - Regional Webinar

#Automated #speech #recognition #The #Singularity #Hidden #Markov #Models
Share
Pikes Peak Computer Society Offers:
 
Speech Recognition: An Historical Perspective on the Business and Technology

  Date and Time

  Location

  Hosts

  Registration


  • Date: 22 Apr 2021
  • Time: 06:00 PM to 08:00 PM
  • All times are (GMT-07:00) US/Mountain
  • Add_To_Calendar_icon Add Event to Calendar
If you are not a robot, please complete the ReCAPTCHA to display virtual attendance info.
  • Contact Event Hosts

  Speakers

Topic:

Speech Recognition: An Historical Perspective on the Business and Technology

Automated speech recognition has gone from clunky demos and wild dreams in the 1950s to a ubiquitous productivity tool today. We’ll sketch this evolution and growth, highlighting some key milestones along the way. In doing so, we’ll explore the who’s, how’s, and why’s of the investments that made this possible, as the technology progressed from analog filters to machine learning. Along the way we’ll explain the basics of audio processing, phonetics, morphology, syntax, semantics, and pragmatics – fancy words linguists use to break down the amazing process of moving concepts and knowledge of the world from one brain to another via language. We’ll even do a shallow dive into the Hidden Markov Models that underlie most of today’s speech recognition systems, and the art of designing dialogs that compensate for their shortcomings. Finally, we’ll touch on the debate about how close we really are to the holy grail of “natural language understanding”.  Is The Singularity™ right around the corner or not?

Biography:

Mark Holthouse is a mostly retired engineer, manager, entrepreneur and high school teacher currently keeping busy by volunteering as a TA in high school computer science classrooms, and recruiting and mentoring other industry professionals to do the same.  He started his engineering career building and testing advanced (for the time) defense systems before switching to the commercial sector to deploy and operate the first pre-internet home banking system for Chemical Bank in New York using Atari 400s. He then started an IVR software company in his dining room, built it to $10M in profitable sales, and sold it to invest in a speech recognition software venture.  As COO of SpeechWorks, he drove speech rec application development and deployment for leading US and international companies, leading to a successful IPO in 2000. He capped his working life teaching math and science in a public high school where he built a complete computer science and engineering program that serves over half of the school’s students, and is now led by a one of his former students.
 
Mark earned a BS CSE from MIT and an MBA from Boston University, and has been active in the IEEE since publishing in Computer Magazine in 1976.  He’s also a member of the ASEE, Sigma Xi, Tau Beta Pi, and Eta Kappa Nu. He was awarded the 2010 MIT Inspirational Teacher Award, and is a Level 4 USA Hockey coach. But, he is most proud of his two engineer daughters, one busily automating factories for Chicago-based engineering services firm DMC, and the other serving as an MC-130 Electronic Warfare Officer in USAF Special Ops. 

Address:United States




Agenda

Automated speech recognition has gone from clunky demos and wild dreams in the 1950s to a ubiquitous productivity tool today. We’ll sketch this evolution and growth, highlighting some key milestones along the way. In doing so, we’ll explore the who’s, how’s, and why’s of the investments that made this possible, as the technology progressed from analog filters to machine learning. Along the way we’ll explain the basics of audio processing, phonetics, morphology, syntax, semantics, and pragmatics – fancy words linguists use to break down the amazing process of moving concepts and knowledge of the world from one brain to another via language. We’ll even do a shallow dive into the Hidden Markov Models that underlie most of today’s speech recognition systems, and the art of designing dialogs that compensate for their shortcomings. Finally, we’ll touch on the debate about how close we really are to the holy grail of “natural language understanding”.  Is The Singularity™ right around the corner or not?