Automated speaker identification (SID) is a crucial step for the personalization of a wide range of speech-enabled services. Typical SID systems use a symmetric enrollment-verification framework with a single model to derive embeddings both offline for voice profiles extracted from enrollment utterances, and online from runtime utterances. Due to the distinct circumstances of enrollment and runtime, such as different computation and latency constraints, several applications would benefit from an asymmetric enrollment-verification framework that uses different models for enrollment and runtime embedding generation. To support this asymmetric SID where each of the two models can be updated independently, we propose using a lightweight neural network to map the embeddings from the two independent models to a shared speaker embedding space. Our results show that this approach significantly outperforms cosine scoring in a shared speaker logit space for models that were trained with a contrastive loss on large datasets with many speaker identities. This proposed Neural Embedding Speaker Space Alignment (NESSA) combined with an asymmetric update of only one of the models delivers at least 60% of the performance gain achieved by updating both models in the standard symmetric SID approach.

Alexa wants to help your child learn how to read. With Amazon's new Reading Sidekick, kids can say "Alexa, let's read," to an Amazon Kids-enabled Echo device or the Amazon Kids app on a tablet and the artificial intelligence-powered assistant will take turns reading with them. An Amazon Kids subscription ($2.99 monthly) is required. Kids can choose from hundreds of physical and digital books that are supported, with more being added monthly. After asking Alexa to read with them, the AI assistant will ask how much do they want to read: a little, a lot, or taking turns.

As good as it is to read with your kids, you might not always be there when they want to open a book. Amazon thinks it can fill in that gap, though. It just rolled out a long-teased Reading Sidekick feature that uses an Echo Kids device to help your kids read aloud on their own time. Children just have to tell Alexa "let's read" to take turns reading supported books, whether they're digital or physical. Your young ones won't always have to wait for you, in other words.

So, you recently bought or were gifted an Echo, Echo Dot, or another Echo device, and it's sitting in your kitchen, silently awaiting your next order. Before you can ask your Alexa-powered Echo to play your favorite Spotify playlist or to turn on your living room lights, you'll need to tweak a few key settings. Get the scoop on how to train Alexa to recognize your voice, keep her from letting just anyone buy stuff on Amazon, tell her where you live and work, and more. As soon as your new Echo is up and running, Alexa can start answering your questions and doing your bidding. That said, it's a good idea to help Alexa get accustomed to your voice as soon as possible.

Scribe Healthcare Technologies Inc. has released a revolutionary speech recognition software called Scribe Interactive. Since physicians and healthcare personnel need to be adaptable to cost efficient ways of practicing healthcare, Scribe Interactive is one tool that can immediately generate transcription layout from a dictation. With Scribe Interactive, the built in Scribe's M*Modal's Speech Recognition Engine leverages existing voice profiles to accomplish this. Even without a recognized voice profile, Scribe Interactive allows users to create verbal snippets for efficiency. There are many added features to this tool.