10月1日 的 Maise 應用程式分析

Maise is an open-source android speech engine designed to provide a powerful and flexible platform for speech synthesis on the edge. To achieve this goal, Maise utilizes the Onnx Runtime to run multiple bleeding edge models to generate high quality speech. For text processing, Open Phonemizer is used to convert text into phonemes, which are then fed into the two speech synthesis models (Kokoro and Kitten TTS) to generate the final audio output. Maise is implemented as a android system speech service, allowing it to work out of the box with any android application that uses the system text-to-speech service.