This project is to analyze the similarity between two audios.
Depends on fpcalc linux extension
sudo wget https://github.com/acoustid/chromaprint/releases/download/v1.4.3/chromaprint-fpcalc-1.4.3-linux-x86_64.tar.gz
sudo tar -zxvf chromaprint-fpcalc-1.4.3-linux-x86_64.tar.gz
sudo mv /usr/local/chromaprint-fpcalc-1.4.3-linux-x86_64/fpcalc /usr/local/bin
fpcalc -version
fpcalc version 1.4.3
first, _ := fingerprint.GetFingerPrintInfo("first.mp3")
second, _ := fingerprint.GetFingerPrintInfo("second.mp3")
similarity, _, _ := fingerprint.Correlate(first.Fingerprint, second.Fingerprint)
fmt.Println(similarity)
go test -test.bench="BenchmarkSongFpCompare" -benchmem -benchtime=30s -cpuprofile profile_cpu.out
go tool pprof -svg profile_cpu.out > profile_cpu.svg
Gochroma can only match exactly
- nay-kang Provide python script