演讲者识别，如面部识别？ [英] Speaker Identification like face identification?

问题描述

我是最近的毕业生，经过几个关于信号处理和机器学习的课程，我喜欢面部识别api！它似乎是一个识别图片中的人或人的神奇工具。如果我理解正确，你只需要
"教"。人们通过提供培训数据然后根据小组继续识别人员。现在，使用相同的原理，有没有办法让人们使用语音片段训练数据并尝试在声音片段中识别人物？ I
知道声音可能会更难以取决于质量和噪音，但这有多难？我对使用这种技术非常感兴趣，并且很想听听牛津计划是否正在努力增加或听取牛津项目如何进行
面部检测，以及如何尝试使用Azure Machine进行应用学会自己做点什么。希望收到回复！

I'm a recent grad and after a couple of classes dealing with signal processing and machine learning, I love the face identification api! It seems like an amazing tool for identifying a person or people in a picture. If I understand correctly you simply "teach" it people by giving it training data and it then proceeds to id people based on the group. Now, using that same principle, is there a way to have a people set of training data with voice clips and try to id the people in the sound clip? I know sound can be more difficult depending on quality and noise but how hard would this be? I'm very interested in using that kind of tech and would love to hear if that is something Project Oxford is working on adding or hearing how project oxford did the face detection and how I could try to apply that using Azure Machine learning to do something on my own. Hope to hear back!

推荐答案

嗨凯文

很棒，你对这一切感兴趣！保持。牛津计划包括称为"说话人识别API"的东西。这可能会帮助你完成你想要完成的任务。

Awseome, that you are interested in all of this! Keep it up. Project Oxford includes something called the "Speaker Recognition API" that might help you with the task you are trying to achieve.

你可以试试这里。

完整的api文档和api测试控制台是
这里。

The full api documentation and the api testing console is here.

希望这会有所帮助！

Mert

这篇关于演讲者识别，如面部识别？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！