Microsoft Upgrades Azure AI Speech Service with 9 More Realistic AI Voices
-
Microsoft has recently made significant upgrades to its Azure AI Speech Service, introducing 9 new and more realistic AI voices designed to provide users with a more natural and immersive conversational experience. The core of this upgrade is the introduction of zero-shot learning (Zero-shot) text-to-speech (TTS) models, which not only improve the naturalness of synthesized speech but also better mimic the characteristics of the prompt voice.
The Personal Voice feature of Azure AI Speech Service, powered by these new models, makes the creation of personalized voices quick and easy. Users only need to provide a short voice sample to generate an AI voice that mimics their unique vocal characteristics in just a few seconds. This highly natural voice output is not only smooth and natural but also accurately captures the subtle nuances of human speech, such as tone, rhythm, and emotional expression, greatly enhancing the vividness and realism of synthesized speech. Experience the update: https://ai.azure.com/explore/speech
Microsoft's service supports over 400 neural voices, covering more than 140 languages and regions, making text-to-speech conversion not only fast but also barrier-free. Additionally, through the custom neural voice feature, users can easily create unique brand voices for their businesses, whether for personalized applications or cross-language dubbing needs.
The application scenarios for Azure AI Speech Service are very extensive, including personalized voice assistants, immersive gaming experiences, multilingual dubbing, media and entertainment content creation, and speech translation. These new Zero-shot TTS models are not only suitable for personalized applications but also provide support for scenarios requiring real-time interaction. Microsoft has particularly emphasized the responsible use of AI, implementing strict guidelines and access controls for the use of the Zero-shot TTS model to ensure the responsible deployment and application of the technology, safeguarding the rights of individuals and society.
Additionally, Microsoft has released 9 AI voices optimized for conversation, covering multiple languages to provide users with more choices and diversity. These voices sound more natural and engaging when reading dialogues and informal texts, even including interjections such as laughter and filler pauses, adding a human touch to virtual conversations.
The new Generally Available (GA) voices include various language options such as English, German, French, and Chinese, all optimized for conversational styles. This expands Microsoft's capability to deliver content in 91 languages and their variants, reflecting the company's strong commitment to overcoming language barriers and promoting more inclusive and accessible global communication. Users can view detailed introductions and demo effects through Azure AI Speech Service's online voice gallery, and further explore these new features via update experience links. Microsoft's upgrade will undoubtedly provide richer and more authentic voice interaction experiences for various business scenarios.