编程知识 cdmana.com

HUAWEI cloud micro authentication DIY small experiment, uncover the shaking of "tiktok" technical mystery

in recent years , Tiktok 、 The rise of fast video and short video software has made some Kwai song songs popular overnight . It's not just the heat of pop music , It also stirs up the hearts of many music enthusiasts , But when you meet your favorite music, you can't find the exact resources , The other “ A barrier to the heart ” And that's what happened ! Fortunately ,“ Listening to songs and recognizing music ” The function of the music platform was quickly launched , It not only provides the most convenient search way for the majority of music lovers , It can also help users locate specific songs quickly through efficient voiceprint recognition and matching database .

today , I'd like to introduce an introduction about “ Listening to songs and recognizing music ” Small experiment of , It can be used by Huawei cloud services and Huawei cloud college micro certification experimental platform , Manual DIY, Tiktok video “ Listening to songs and recognizing music ” The whole process of Technology .

Micro authentication 《 Tiktok recognition of background video name 》, For short video and music enthusiasts , And social groups interested in the direction of AI voiceprint recognition , Combining cloud database and Python, With the technology of speech processing, the background song name can be recognized quickly , Can take you to learn Huawei cloud database RDS And speech processing related concepts and practice , master RDS And Python Joint deployment of experiments to improve the corresponding technical operation ability .

The experiment passed Python Huawei cloud database RDS service , Convert short video to audio format and extract voiceprint data , And compared with the voiceprint features in the voiceprint database , So as to obtain the background song name , Realize to Recognition of background songs in small tiktok videos . Does that sound very interesting? There are dry goods ? Keep learning with me !

Tiktok “ Listening to songs and recognizing music ” see “ Speech processing technology ”

Previously recognized song names , We all distinguish lyrics by ear , Then through the search engine query lyrics to get the corresponding song name . But because of the different tone and singing skills of the song , It usually leads to errors in the identification of lyrics , As a result, it is impossible to find the corresponding song accurately .

Tiktok “ Listening to songs and recognizing music ” And “ Speech processing technology ”

In this case , We use voice processing technology to solve this problem

Solution - Speech processing technology

Voice signal processing (Speech Signal Processing) Speech processing for short , It is used to study the process of speech production 、 Statistical characteristics of speech signals 、 Automatic speech recognition 、 Machine synthesis, speech perception and other processing technologies .

Because modern speech processing technology is based on digital computing , And with the help of microprocessors 、 Signal processor or general purpose computer , Therefore, it is also called digital speech signal processing .

 

Speech processing is also called speech signal processing , It is a process used to study speech production . Through voice processing technology , We can improve “ Tiktok recognition of background video name ” The accuracy of . Simply speaking , We match the extracted speech features with the existing database , And then through the comparison with speech features , Get the corresponding song name .
    Before speech processing , First of all, we need to do voiceprint recognition . Voiceprint recognition as a kind of biometric recognition , It can also be called speaker recognition , The feature information of voiceprint is extracted from the speech signal sent by the speaker . There are usually two stages , That is, voice registration and voice verification .

Speech processing “ Voiceprint registration ” and “ Voiceprint verification ”

 

When we register the voice , It can be used to extract speech features , Get the voiceprint and import it into the database to establish the corresponding voiceprint database . After getting the voice file , The voiceprint features are obtained by preprocessing , After comparing with the voiceprint database , The best recognition result is obtained .

 

 “ Tiktok recognition of background video name ” Scheme structure

In understanding “ speech recognition ” After technical background , How to proceed “ Tiktok recognition of background video name ” Well ?

Simply speaking , We can install Python3, Create Huawei cloud database that meets the experimental requirements RDS, Configuration correspondence Python3 The required environment variables and related components ensure that Python3 Normal use , Create Huawei cloud database RDS And corresponding elastic public network IP, Make sure Python3 Be able to work with Huawei cloud database RDS Data connection , Can finish “ Tiktok recognition of background video name ” The task of .

“ Tiktok recognition of background video name ” Scheme structure

Want to go through this simple DIY A little experiment , Easy to master “ Tiktok recognition of background video name ” The mystery behind it ? Want to understand cloud database systematically , Study Python Do you ? Come to Huawei cloud College , One stop online learning 、 Experiment and examination ! Obtain official certification certificate of Huawei cloud micro certification , It can also add energy to the workplace upgrade !https://edu.huaweicloud.com/certifications/728fe3c3f88b44689775aa3d2f356246

版权声明
本文为[Edu of Huawei cloud College]所创,转载请带上原文链接,感谢

Scroll to Top