WaveFake Mixed Voice Dataset (2s & 4s) — LJSpeech + EnvSDD
收藏官方服务：
资源简介：
Mixed audio dataset for deepfake voice detection using WaveFake sources. Real voice: LJSpeech (English). Fake voice: WaveFake (LJSpeech-based vocoders). Background: EnvSDD (real/fake environmental sounds). 4 categories × 3 volume levels × 1000 files = 12,000 WAV files each for 2s and 4s versions.
提供机构：
Zenodo创建时间：
2026-06-06
