Amazon shows the novel AI technology this week, including taking a more conversational voice model to better compete with such things as the advanced Gemini’s Voice mode live and OpenAI and update the model that can generate video.
The novel Nova Sonic voice model supports real -time speech processing and generating AI voice for conversation applications, Amazon says. Nova Sonic uses the “united model architecture”, which Amazon claims, is better than other approaches that combine separate models for speech recognition, speech conversion to text, generating answers, and then text-audio. Amazon says that Nova Sonic can also better detect someone’s tone and provide more natural answers.
Nova Sonic is available to try the Amazon Bedrock Developer platform, and the company claims that it can be used to create things such as customer service bots or build AI agents for travel, education, healthcare and many other industries. Nova Sonic’s “Components” are already used in the novel assistant Alexa Plus Amazon, Amazon’s Rohit Pragad, SVP and the main scientist of Aga, he said TechCrunch.
As for the video, Amazon announced Nova Reel 1.1Which, according to the company, ensures improvement of quality and delays compared to 1.0. It can now also store coherent styles in many 6-second scenes cut to a full movie of up to two minutes.