Zhipu AI announces that its end-to-end emotional voice technology has officially launched on the Zhipu Qinyan platform and is now available to all users. This technology overcomes the limitations of traditional text-to-speech (TTS) technology, capable of deeply understanding contextual nuances and generating emotionally rich natural dialogues. This marks the evolution of Zhipu AI's speech synthesis technology from simple text reading to an AI that can express genuine emotions.

Alibaba's latest CosyVoice speech synthesis model and SenseVoice speech recognition model together form the FunAudioLLM framework, aimed at enhancing human-computer interaction experience. CosyVoice, with its realistic voice generation capability, can mimic voices of different genders, ages, and personalities, adding emotions and styles while even simulating natural features such as laughter, coughing, and breathing. SenseVoice focuses on high-precision multilingual speech recognition, emotion detection, and audio event detection, supporting over 50 languages.

iFLYTEK will release the iFLYTEK Spark V4.0 on June 27 this year, which will showcase the latest achievements in end-to-end intelligent voice technology. In addition, iFLYTEK plans to launch its end-to-end speech technology modeling solutions in the next three quarters. Furthermore, iFLYTEK will provide users with one-sentence replication capabilities that GPT-4 does not possess.

Microsoft's latest technology, Personal Voice, generates AI voice that perfectly matches the user's voice in one minute. It supports output in 100 languages, including Chinese, Spanish, German, and more. It is widely used in voice assistants, gaming, entertainment, and media. In collaboration with Haier, it brings personalized voices into smart speakers and home appliance products. Prior to launch, it committed to responsible usage, offering limited registration, and adding watermarks to ensure compliance and transparency.

AI can detect type 2 diabetes by merely listening to a patient's speech for 6-10 seconds. Research shows an accuracy rate of 89% in females and 86% in males. The study analyzed over 18,000 recordings using voice technology, successfully identifying differences in voice between diabetic and non-diabetic patients.