4.技术应用上述技术的落地,使得声纹识别技术可以在原本的许多“无数据,不可用”的场景中落地应用。在无人工标注数据的情况下,通过模型的自动标签与迭代,快速支持多款不同需求的智能音箱产品上线;基于远场交互的声纹识别技术,也在其它智能家居场景上成功落地应用。 5. 参考文献[1] S. Zheng, G. Liu, H. Suo, and Y. Lei,“Autoencoder-based semi-supervised curriculum learning for out-of-domainspeaker verification,” in INTERSPEECH2019 – 20th Annual Conference of the International SpeechCommunication Association, September 15-19, Graz, Austria, Proceedings, 2019.[2]S. Zheng, G. Liu, H. Suo, and Y. Lei, “TowardsA Fault-tolerant Speaker Verification System,” inINTERSPEECH 2019 – 20thAnnual Conference of theInternational Speech Communication Association, September 15-19, Graz, Austria,Proceedings, 2019.