Speech Translation for Everyone – Breaking Down the Language Barriers – Will Lewis (Microsoft Research)

April 17, 2015 all-day

Fifty years ago Star Trek had the Universal Translator. Thirty-five years ago we were introduced to the Babel Fish in The Hitchhiker’s Guide to the Galaxy. Decades later, is reality finally catching up to science fiction? Given the enormous strides made in speech recognition and machine translation over the last decade, is all that is required just a matter of chaining together speech recognition and machine translation? In fact, recent improvements in machine learning, as well as increased computing capacity and large supplies of data, have tipped the balance, making it possible to build a near real-time conversation translator, such as the one we have built in Skype Translator. In the Skype Translator project we set ourselves an ambitious goal – to enable successful open-domain conversations between Skype users in different parts of the world, speaking different languages. As one might imagine, putting together two error-prone technologies such as speech recognition and machine translation raises some unique challenges. In this talk, I will share what we have learned over the course of the Skype Translator project. I will discuss what we are doing to bridge the gap between ASR and MT, how we are adapting our ASR and MT systems to the real world challenges presented by open-domain conversational scenarios, and what it takes to get this technology into the hands of real users. I will also touch upon some of the open issues and challenges we still face. I will end with a demo of the Skype Translator with a remote non-English speaker.

William Lewis is Principal Technical Program Manager with the Microsoft Translator team at Microsoft Research. He has led the Microsoft Translator team’s efforts to build Machine Translation engines for a variety of the world’s languages, and has been working with the Translator team to build the Skype Translator. Before joining Microsoft, Will was Assistant Professor and founding faculty for the Computational Linguistics Master’s Program at the University of Washington, where he continues to hold an Affiliate Appointment, and teaches classes on Natural Language Processing. Before that, he was faculty at CSU Fresno, where he helped found the Computational Linguistic and Cognitive Science Programs at the university. He received his Bachelor’s degree in Linguistics from the University of California Davis and a Master’s and Doctorate in Linguistics, with an emphasis in Computational Linguistics, from the University of Arizona in Tucson.

Center for Language and Speech Processing