最新发布!查看 MagicData 成品数据集
语音复刻大模型高品质数据集——东南亚

语音复刻大模型高品质数据集——东南亚

语种

他加禄语-菲律宾,马来语-马来西亚,泰语-泰国

数据风格

对话式&朗读式

音频格式

PCM

采样率

48kHz

比特率

16 bits

声道

1

人数

5,232

时长

2,425

As a new trend of AIGC, zero-shot speech synthesis has wide-ranging applications, including voice assistants, audiobooks, video game character voices, creating podcast, and real-time voice changer, among others. Magic Data has proactively designed and developed the "High-Quality Dataset for Voice Replication," which boasts a 48kHz high sampling rate, tens of thousands of contributors with diverse content, making it an excellent resource for achieving zero-shot speech synthesis.

没有找到想要的数据?

联系我们

数据集推荐

MDT-AF027 葡萄牙语对话音频数据集

试听音频

MDT-BE010 中文女声客服TTS数据集

试听音频

MDT-BD003 中文女声情感TTS数据集

试听音频

MDT-AB010 中英混合手机朗读音频数据集—命令控制

试听音频

MDT-AF076 云南方言对话音频数据集

试听音频

MDT-AJ039 日语双工自然对话数据集

本数据集通过高保真独立音轨完整收录日常对话中的打断、重叠发言、语调起伏及情感停顿等自然交互特征。所有对话均经多说话人分类标注,覆盖多样化场景,为AI模型提供复杂日语敬语体系、口语省略表达及上下文逻辑的深度学习资源。

Magic Data 数据优势有哪些?

合规性

符合ISO/IEC 27001和ISO/IEC 27701:2019标准认证

多维度

音频、文本、图像、音视频多模态数据

覆盖性

涵盖多领域的对话式、朗读式及自发式数据

高精度

人机协同高精度标注

TOP
联系我们