jianchang512/ChatTTS-ui
ChatTTS Ui
ChatTTS-ui es una interfaz web y un wrapper de API para el modelo ChatTTS, diseñado para convertir texto escrito y entradas de lenguaje mixto en audio hablado. Funciona como un panel de síntesis de voz por IA y un generador programático para crear salidas de voz naturalistas.
El proyecto se centra en la creación de perfiles de voz personalizados y el control de matices del habla. Permite mantener características de voz consistentes utilizando valores semilla y archivos de datos, mientras proporciona controles para el tono, la risa y las pausas a través de prompts de comportamiento y parámetros de muestreo.
El sistema incluye una arquitectura cliente-servidor que maneja el procesamiento de audio asíncrono y proporciona una interfaz programática para la integración con aplicaciones externas. Gestiona perfiles de voz y configuraciones de audio a través de una interfaz con gestión de estado para asegurar una síntesis consistente.
Features
- Text-to-Speech Synthesis - Converts written text and mixed language input into natural-sounding spoken audio.
- API Speech Synthesizers - Provides a programmatic interface for external applications to convert text into synthesized audio files.
- Voice Synthesis - Generates natural sounding speech with precise control over tone and pauses using the ChatTTS model.
- Text-to-Speech Conversions - Converts written text into spoken audio using AI language models for various projects.
- Text-to-Speech Integrations - Exposes ChatTTS model functionality as a set of API endpoints for external application integration.
- Vocal Nuance Controllers - Controls non-verbal cues such as laughter and pauses to enhance the naturalness of the speech.
- Voice Profile Management - Loads and manages speaker characteristics through data files and seed values for consistent output.
- Prompt-Based Audio Generation - Leverages large language models to produce naturalistic spoken audio based on configurable text prompts.
- Generative Audio APIs - Provides programmatic interfaces for triggering the generation of synthesized speech from external scripts.
- Prompt-Driven Parameter Synthesis - Allows fine-tuning of voice nuance and tone using behavioral prompts and sampling parameters.
- Voice Identity Seeds - Maintains consistent speaker characteristics using specific seed values during audio generation.
- Multilingual Synthesis - Transforms mixed language text into spoken audio through a unified synthesis interface.
- Synthesis Control Dashboards - Provides a visual workspace for adjusting speech nuance, sampling parameters, and voice profiles.
- Asynchronous Request Processing - Implements asynchronous processing to manage long-running speech synthesis tasks between the UI and backend.
- REST API Integrations - Connects the web interface to the synthesis server via RESTful HTTP requests.
- Web Interfaces - A simple local web interface with API support.
