Added audio length (in seconds) to database. Added relative confidence value. Created branch to avoid pull request #148 issues (more commits)

worldveil / dejavu

Audio fingerprinting and recognition in Python

MIT License

6.36k stars 1.43k forks source link

Added audio length (in seconds) to database. Added relative confidence value. Created branch to avoid pull request #148 issues (more commits) #197

Open JPery opened 5 years ago

JPery commented 5 years ago

This pull request is made in order to avoid the extra commits present in #148 so it can be closed

gitteraz commented 4 years ago

Hi @JPery How do i use the relative confidence of the returned song to know if it is the correct song?

JPery commented 4 years ago

As you can see in the attached graphic at https://github.com/worldveil/dejavu/pull/148#issuecomment-403382614 better than 1.2 should be enough to validate the identification. It depends on your use case, the lower the value the more FPs you can get. In my use case I use a value of 2.

gitteraz commented 4 years ago

Thank you for your answer. In the graphic the Y Axis represents “relative confidence” in percentage right? So you mean you use a value of 2%?

Also can I ask if that was always accurate for you and if you think it can be always accurate on a set of more than 50.000 fingerprinted songs.

Thank you