Introducing the Personal Voice Feature by Apple
The Personal Voice feature, introduced in May 2023 and made available on iOS 17 in September 2023, is a powerful tool for people facing the risk of losing their ability to speak, such as those with a recent diagnosis of amyotrophic lateral sclerosis (ALS) or other conditions that can progressively impact their speaking ability.
How It Works
Personal Voice is a tool that creates a synthesized voice for users to speak in FaceTime, phone calls, assistive communication apps, and in-person conversations. To create their personalized voice, users read aloud a set of text prompts to record 150 sentences on their latest iPhone, iPad, or Mac. The voice audio is then tuned with machine learning techniques overnight directly on the device while it is charging, locked and connected to Wi-Fi.
Features
The Personal Voice system includes three machine learning approaches: the Personal Voice TTS system, voice model pretraining and fine-tuning, and on-device speech recording enhancement. Through these approaches, the system provides users with the ability to create their own personalized voice, maintain privacy and security, and achieve a better overall quality of voice output.
The Personal Voice TTS system is a neural TTS system that converts written text to speech output. It includes three major components: text processing, acoustic model, and vocoder model. The system fine-tunes both the acoustic and vocoder models on-device for better voice quality and similarity to the target speaker’s voice.
Voice Model Pretraining and Fine-Tuning consists of modified FastSpeech2-based acoustic and WaveRNN-based vocoder models. The system uses a general pretraining and fine-tuning strategy, and all fine-tuning occurs on the user’s Apple device for faster and more satisfying conversation experience.
On-Device Speech Recording Enhancement allows users to record their voice samples and applies speech augmentation to achieve the best voice quality. This approach significantly improves the quality of the generated voice, especially with real-world recorded data.
Conclusion
The Personal Voice feature by Apple provides a powerful and accessible tool for users to create their own voice overnight and use real-time speech synthesis to communicate with others. This innovative AI-powered technology is a game-changer for those at risk of losing their ability to speak, such as those with ALS. Give it a try and experience the benefits of Personal Voice for yourself!