Amazon Developer Blogs

Amazon Developer Blogs

Showing posts tagged with Alexa Science

July 02, 2018

Larry Hardesty

The deadline for the 2018 Amazon Research Awards, a grant program that provides up to $80,000 in funding and $20,000 in Amazon Web Services Promotional Credits to academic researchers investigating topics related to machine learning, is September 15. 

[Read More]

June 08, 2018

Young-Bum Kim

Amazon science leader Young-Bum Kim describes the HypRank neural model, which performs the second step in the two-step process that Alexa uses to select the best skill to handle a given utterance.

[Read More]

June 07, 2018

Young-Bum Kim

Amazon science leader Young-Bum Kim describes the Shortlister neural model, which performs the first step in the two-step process that Alexa uses to select the best skill to handle a given utterance.

[Read More]

June 01, 2018

Lambert Mathias

At a computational linguistics conference that begins this weekend, Amazon scientists will present a new, more sophisticated semantic-representation language. Data annotated in the language should enable Alexa skills to handle more sophisticated conversational interactions. 

[Read More]

May 29, 2018

Penny Karanasou

In a paper that will be presented next week at a North American computational linguistics conference, Amazon scientists propose an approach for quickly bootstrapping machine-learning models in new languages, with the goal of bringing Alexa to new customers around the world more quickly.

[Read More]

May 24, 2018

Angeliki Metallinou

At a forthcoming Computational Linguistics Conference, Amazon scientists will present how they use transfer- and deep-learning techniques to help them scale Alexa's natural-language-understanding capabilities faster, with less data.

[Read More]

May 11, 2018

Jun Yang

Amazon scientist Jun Yang recently published and presented her work on a multilayer echo cancellation and voice enhancement processing system that can significantly improve smart speaker automatic speech recognition and full duplex voice communication performance.

[Read More]

May 04, 2018

Arpit Mittal

Researchers from Amazon and the University of Sheffield are making a large-scale fact extraction dataset publicly available for the first time, and organizing a machine-learning competition for academic and industry researchers to tackle the problem. 

[Read More]

April 25, 2018

Ruhi Sarikaya

Alexa scientists and engineers are on a multi-year journey to fundamentally change human-computer interaction. Skills arbitration, context carryover and a new memory feature are early instances of a class of work to make engaging with Alexa more friction-free. 

[Read More]

April 12, 2018

Minhua Wu

The first step in providing a great customer experience with an Alexa-enabled device is wake word detection. Next week, at the IEEE International Conference on Acoustics, Speech and Signal Processing, Amazon scientists will present two new techniques that improve wake word performance.

[Read More]