Tuplemax Loss for Language Identification
Tuplemax Loss for Language Identification
复制标题
用于语言识别的 Tuplemax 损失
DOI：
10.1109/icassp.2019.8683313
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
I. López
中科院分区：
文献类型：
--
作者：
Li Wan;Prashant Sridhar;Yang Yu;Quan Wang;I. López
在语言识别任务的许多场景中，用户将指定他/她可以说的一小组语言，而不是所有可能的语言的一大组。我们希望通过将常用的Softmax损失函数替换为一种名为tuplemax Lost的新损失函数，将这些先验知识建模为我们训练神经网络的方式。事实上，在北美推出的典型语言识别系统约有95%的用户会说不超过两种语言。使用tuplemax损失率，我们的系统获得了2.33%的错误率，比标准Softmax损失法3.85%的错误率相对提高了39.4%。
In many scenarios of a language identification task, the user will specify a small set of languages which he/she can speak instead of a large set of all possible languages. We want to model such prior knowledge into the way we train our neural networks, by replacing the commonly used softmax loss function with a novel loss function named tuplemax loss. As a matter of fact, a typical language identification system launched in North America has about 95% users who could speak no more than two languages. Using the tuplemax loss, our system achieved a 2.33% error rate, which is a relative 39.4% improvement over the 3.85% error rate of standard softmax loss method.