谷歌发布了语音分离数据集与深度学习模型。语音分离数据集,全名为The Free Universal Sound Separation (FUSS) Dataset 。这个数据集将在DCASE2020 Challenge Task 4: Sound Event Detection and Separation in Domestic Environments作为官方数据集。所有的数据格式是uncompressed PCM 16 bit, 16 kHz, mono audio files。此外,也发布了一个baseline的模型。
github地址:https://github.com/google-research/sound-separation
数据集介绍:https://github.com/google-research/sound-separation/blob/master/datasets/fuss/FUSS_license_doc/README.md
参考文章:https://arxiv.org/abs/1810.04826
《VoiceFilter: Targeted Voice Separation by Speaker-Conditioned Spectrogram Masking》
模型地址:https://github.com/google-research/sound-separation/blob/master/models/dcase2020_fuss_baseline/README.md
