New algo to identify music using audio-fingerprinting

In Progress Posted 5 years ago Paid on delivery
In Progress Paid on delivery

Seeking help in making algorithmic changes to an existing open-source python 2.7. component dejavu -- which is generating an audio-fingerprint for MP3 files/data-snippets. The audio-fingerprint algorithm is e.g. used by shazam to identify unknown audio/music recordings.

In our scenario we have access to 10 seconds MP3 content of a music file (always the beginning) and we want to create a DB of audio-fingerprints from these 10 seconds in order to identify music for which we don't have ID3 data and only an un-descriptive filename.

The (dejavu) audio-fingerprinting algorithm is able to detect music even with a relative high signal to noise ratio, but this is not the scenario that we are dealing with -- and therefore we don't need the large redundancy from the generated data coming from that approach.

We are seeking a freelancer who is creative enough to imagine a very similar algorithm that is better adapted to our scenario and that is also applicable to video identification (i.e. from extracted audio)

The actual extraction of MP3 from audio/video files is already done. We only need an algorithm that is extracting identifiable signals from 3,4 or 5 secs of MP3 audio in order to have a very high confidence level that we have identified the right music in a large DB of audio-fingerprints.

The product deliverable should be provided in Python 3.x.

The dejavu code is in [login to view URL] -- additional docu on audio-fingerprinting can be found in: [login to view URL]

Algorithm Audio Processing Database Development Python Software Architecture

Project ID: #18554723

About the project

3 proposals Remote project Active 5 years ago

3 freelancers are bidding on average $246 for this job

KGeorgy

Hello, I am pleasure with your job for New algo to identify music using audio-fingerprinting. Thank you for the job posting. It’s a pleasure to meet you. I’d really like to work with you on this one if possible! I More

$155 USD in 3 days
(18 Reviews)
5.3
anamaria1979

he (dejavu) audio-fingerprinting algorithm is able to detect music even with a relative high signal to noise ratio, but this is not the scenario that we are dealing with -- and therefore we don't need the large redunda More

$250 USD in 2 days
(27 Reviews)
5.4
oscarBeta

Hi, your project does arouse my research interest, whereas i came across the webpage explaining the dejavu algorithm for extracting hashed fingerprint from 10s mp3 music segment. I have worked on implementation of r More

$333 USD in 14 days
(0 Reviews)
0.0