论文标题
自动语音验证欺骗检测
Automatic Speech Verification Spoofing Detection
论文作者
论文摘要
自动语音验证（ASV）是根据人的声音确定人身份的技术。虽然可以方便身份验证，但鉴于它是有价值的数字资产的保障，我们应该以最高的系统安全标准为目标。考虑到这一点，我们遵循ASVSPOOF 2019竞赛的设置，以开发出强大而有效的潜在对策。两个指标EER和T-DCF将用于系统评估。
Automatic speech verification (ASV) is the technology to determine the identity of a person based on their voice. While being convenient for identity verification, we should aim for the highest system security standard given that it is the safeguard of valuable digital assets. Bearing this in mind, we follow the setup in ASVSpoof 2019 competition to develop potential countermeasures that are robust and efficient. Two metrics, EER and t-DCF, will be used for system evaluation.
