WARNING: This article may be obsolete
This post was published in 2022-09-06. Obviously, expired content is less useful to users if it has already pasted its expiration date.
This post was published in 2022-09-06. Obviously, expired content is less useful to users if it has already pasted its expiration date.
This article is categorized as "Garbage" . It should NEVER be appeared in your search engine's results.
Audio Matching
🔗 [C7S2_AudioMatching] https://www.audiolabs-erlangen.de/resources/MIR/FMP/C7/C7S2_AudioMatching.html
基本思路就是用DTW计算匹配出一个“绑在时间序列上的match score(反相关的score,数值越低越好,见下图)”,然后找出这个match score曲线的几个最低点即可。
大概就是这个样子
2023年,我来写点对这个项目的代码的看法吧:
这段audio matching代码的准确度其实已经很高了,我觉得是优化到头了,(在现有算法框架的基础上)我实在想不出什么可以优化的方向。
非常准确:原音频匹配,10秒以上
次准确:原音频匹配,5秒。对5秒的音频有一定要求(比如不能是那种拖长音的音频截取,要有一定节奏,这样比较容易配上)
比较不准确:非原音频匹配(其他版本,或者remix),5秒。这个不准确是很正常的,毕竟原来的程序也只有稍稍改进的DTW.
Last Modified in 2023-06-12