Project Overview: Curses
Curses is an innovative software solution designed to convert speech into text captions across a variety of streaming and communication platforms such as OBS (Open Broadcaster Software), VRChat, Twitch chat, and Discord. This project addresses the growing need for accessibility and engagement among users by providing accurate speech-to-text (STT) capabilities.
Key Features
Curses offers a wide range of features to enhance user experience and accessibility:
-
OBS Captions Customization: Users can personalize captions with various colors, fonts, shadows, and background textures. It also includes options for text typing animations, sound effects, particle effects, and the use of CSS for further customization. This allows streamers to create visually engaging and accessible streams.
-
Native OBS Stream Captions: The software integrates seamlessly with OBS to provide native streaming captions, ensuring that all speech is accurately and efficiently translated into text in real-time.
-
Google Fonts Support: With access to over 1000 free fonts, users can customize their OBS captions to fit the style and theme of their streams.
-
Speech to Text Technologies: The project supports several leading STT technologies including Microsoft Azure Cognitive Services, Speechly, Deepgram, and WebSpeechAPI for both Chrome and Edge browsers.
-
Text to Speech (TTS) Technologies: For content creators looking to convert text back into speech, the software supports a range of TTS technologies such as Microsoft Azure, Uberduck, TikTok, Windows API (SAPI), and WebSpeechAPI.
-
VRChat Integration: Curses connects with VRChat through features like KillFrenzy Avatar text and integration with VRChat's chatbox, enhancing user interaction within virtual environments.
-
Twitch Chat Integration: Streamers can integrate popular emotes from 7TV, FFZ, and BTTV into OBS captions, share speech-to-text outputs directly in chat, and use chat messages as a source for captions and TTS.
-
Discord Integration: Users can send their speech-to-text outputs to specified Discord channels, allowing for real-time communication and engagement in their communities.
-
Scene Management in OBS: Streamers can save multiple caption designs and switch between them, either manually or automatically when OBS changes scenes, providing flexibility and efficiency.
Community and Support
Curses boasts a vibrant community that supports users through feature requests, bug reports, and releasing notifications. Users can join the project's Discord server to tap into community resources, receive prompt updates, engage with other users, and access a library of design templates for further customization.
Getting Started with OBS
To begin using Curses with OBS:
-
Download the App: Users can download the latest release from the project's GitHub page or join the Discord server for notifications and updates.
-
Set Up OBS: The app provides an option to copy a link for OBS setup or utilize the "Set Up OBS" button for automatic configuration through the obs-websocket plugin.
-
Create a Browser Source: In OBS, users create a browser source by pasting the provided link, adjusting the window size to match the canvas size of the app, with a default setting of 500x300 for optimal display.
Future Development Plans (Roadmap)
The project aims to expand its functionality with the planned integration of additional STT and TTS technologies, including Vosk for speech-to-text and VoiceVox for text-to-speech, which will allow for further versatility and broadening its accessibility impact.
In summary, Curses is a comprehensive tool that enhances digital communication by breaking down barriers in livestreaming and online interaction through advanced speech-to-text transformation and robust customization features.