Home » News » Tech » Amazon Alexa Can Now Sound Excited, Happy or Disappointed
1-MIN READ

Amazon Alexa Can Now Sound Excited, Happy or Disappointed

Image for Representation

Image for Representation

The new changes in Alexa use Neural TTS technology which results in a more natural-sounding speech, such as mimicking news anchors and radio hosts.

Most of us are used to hearing a robotic voice respond to queries on Amazon's Alexa. However, going by recent developments, that is set to change soon. Amazon has announced that developers can now make Alexa respond to questions from users in the US in varying tones or emotions. These include happy, excited, disappointed, or empathetic tones. Calling the development "new Alexa capabilities", Amazon has stated that it will help create a more natural and intuitive aural experience for users. Amazon has revealed that emotional responses are relevant to skills in gaming and sports categories. Furthermore, according to them, one can also have Alexa respond in a speaking style that is "more suited for a specific type of content."

According to Amazon, the new change will see Alexa's voice mimicking the style followed by TV news anchors and radio hosts when speaking on news among others. Alexa emotions incorporate Neural TTS (NTTS) technology that enables a more natural-sounding speech. It uses Amazon's text-to-speech technology and can have Alexa respond in an excited manner when answering trivia questions or winning a game. Furthermore, they have revealed that US users can now enable 2 different speaking styles, news and music.

In Australia, they can enable an Australia-specific news speaking style as well. According to Amazon, 'blind listening' tests have shown that the news style was seen to be 31 per cent more natural than Alexa's standard voice and the music style 84 per cent more natural. To enable Alexa emotions, one can use SSML tags. One has to simply wrap Alexa's response with the required SSML tag for the required tonality in response, Amazon has revealed on their blog.