Microsoft has disclosed the rollout of seven new AI-created text-to-speech voices for the patrons of its Azure AI Services, following the announcement of three others in US English, along with one in Chinese back in September. This unveiling of additional voices, and updates to previously launched ones is being provided via a sneak peek.
Lifelike Realism for Conversational Platforms
Microsoft aims to improve user interactions with AI voices tailored for conversation. Whether building a chatbot that works on verbal inputs, an oral assistant, or a conversation-driven application, these new voices aimed to bring more authenticity and appeal to interactions. The freshly minted, convincing conversational voices cater perfectly to any utility that necessitates lifelike verbal dialogues. These comprise but are not limited to, chatbots, voice-controlled assistants, gaming platforms, e-teaching applications, entertainment, and more.
Microsoft had previously revealed the process behind creating its AI voices, which are designed to sound highly realistic. The process starts with developing a persona for each voice that portrays a friendly, real individual who is eager to help others and provide practical or appealing guidance. The speech patterns are modeled on a natural conversation, similar to chatting with a friend over coffee. Any spelling, grammar, or punctuation errors have been corrected.
Updated Features for Enhanced Expressivity
The new AI voice features introduced by Microsoft are more natural in their speaking manner. The Xiaoxiao female Mandarin Chinese (China) voice that was previously launched has now been equipped with a more naturally inflected speaking pattern. Additionally, three more voices have undergone updates to incorporate an ‘expressively modulated speech rhythm'.
Currently, Microsoft provides more than 400 neural voices that span over 140 languages and regions through Azure. The company has an exclusive offering for its business clientele in the form of their Custom Neural Voice feature that enables enterprises to construct their distinct digital spokesperson.