Walleclipse / Deep_Speaker-speaker_recognition_system

Keras implementation of ‘’Deep Speaker: an End-to-End Neural Speaker Embedding System‘’ (speaker recognition)
246 stars 81 forks source link

关于threshold问题 #17

Closed MichelleYang2017 closed 5 years ago

MichelleYang2017 commented 5 years ago

您好,又来打扰您了,我看您在选择阈值的时候时从0到1之间选取的,理由是什么呢,相似度按理说不应该时在-1到1这个取值范围吗?

Walleclipse commented 5 years ago

我没有太明白你说的阈值是指什么? cosine-smilarity的取值在 [-1,+1]这个范围内。triplet-loss的取值是大于0的。

MichelleYang2017 commented 5 years ago

谢谢