We are seeking for software solution which will analyze online broadcasting and recognize the composition by comparison of each second to the database.
The software will receive an MP3 file as input once (for each composition), analyze it according to special algorithm, which will create a unique encoding value to each composition.
The software will analyze and compare played composition in real time and provide real time results of artist and name of composition.
Analyze process of the broadcasted stream, will be by comparing played composition to the database. The software will compare every second of the composition, in order to determine the was the composition fully played or not.
Please pay attention, the database may contain data on hundreds of thousands of MP3 files and recognition process (Stream Broadcast recognition) must be especially effective (within a few seconds).
Priority will be given to solution with lowest times per composition.
Software will be able to absorb a number of broadcasts simultaneously and make the detection work in parallel mode.
The software can be divided to two separated software, the first will analyze the mp3 files and create a “fingerprint” for each composition and create a database of “fingerprints”, while the second software will analyze broadcasted stream to the “fingerprint” and determine artist and composition name
Please note in your bid :
1. Hardware requirements you would make them vital and necessary for maximum productivity of the system if there are.
2. Approximate time, you need for our order.
3. Estimated price (final price will be determined together after you get the exact full characterization).
4. Your prior experience in sound recognition systems (if there is)
5. Programming languages which you planning to use.
More accurate characterization and description will be provided after we review your offer and find your company appropriate.
