鸣潮角色TTS数据集

鸣潮角色语音数据集是基于Bert-vits2开源项目制作的，包含了鸣潮游戏1.0至1.3版本的中、日、英、韩全角色语音。它提供了单角色包和完整包以满足不同用户的需求，每个语音文件都有对应的文本标注，方便进行语音识别和训练。数据集中的语音按类别整理，包括战斗语音、带变量语音等，分类清晰，便于查找和使用。用户可以在ModelScope平台直接下载数据集，无需登录。此数据集仅供个人欣赏和学习交流使用，不得用于商业用途或非法活动。鸣潮角色语音数据集旨在支持AI语音合成和识别技术的研究与开发，为用户提供了一个丰富的多角色、多语种的语音资源库。The Wuthering Waves character voice dataset is developed based on the open-source Bert-vits2 project. It covers full voice lines of all in-game characters from versions 1.0 to 1.3 of Wuthering Waves, in Chinese, Japanese, English and Korean. The dataset provides both individual character packs and a complete pack to meet the requirements of diverse users. Each voice file is paired with corresponding text annotations, which facilitates speech recognition and model training. The voices in the dataset are clearly categorized into types, including combat voice lines, variable dialogues and more, making them convenient for searching and utilization. Users can directly download the dataset on the ModelScope platform without logging in. This dataset is only allowed for personal appreciation, learning and communication purposes, and shall not be used for commercial purposes or illegal activities. The Wuthering Waves character voice dataset is intended to support the research and development of AI speech synthesis and recognition technologies, providing users with a rich multi-character, multilingual voice resource library.

搜集汇总

背景与挑战

背景概述

该数据集是基于《鸣潮》（Wuthering Waves）游戏角色的文本到语音（TTS）数据集，由Genius-Society提供集成代码，支持中文、日语、英语和韩语多语言处理。它旨在简化TTS开发流程，通过自动化数据搜索、下载和预处理，方便Python开发者快速集成和使用，适用于语音合成相关应用。