mohammed alshaweesh wrote:
Give me the code or a project
Are you kidding? Audio wave from microphone — of course, but comparison of audio…
How can you be so naive? This is probably beyond the contemporary level of computing. Even is some successful research exist (which I can believe), you hardly can find something freely accessible and really working. How can you imagine the criteria for such comparison? The same audio fragment which can be easily identified by human beings as related or similar can come at different tempo, noises, timbre, etc. For example, the same musical piece can be perform using different instruments and voices. A qualified human being can recognize the similarity, but a program? Do you require such things? Comparison criteria can be very complex, and you did not mention any.
Even though such a complex thing as speech recognition almost becoming successful, the problem of comparison is very different. One problem is the social one: speech recognition is a subject of extensive research, but "abstract" comparison of audio — hardly. What would be the purpose to spend a lot of resources in this direction? Audio Internet search? Everyone understands we are to far from solving such problems, so such research hardly could be commercialized in observable future. So even some research exist, you can hardly expect much financing in this field.
The last thing I tries were the system of recognition of musical tones from the digital audio source. Such programs attempt to create MIDI our of raw audio. Even though this problem is one of the easiest in the field of audio recognition (after all, instruments tuners are usual and are very effective and precise), the accuracy of recognition was quite poor, pretty far from being useful. I someone tells me this problem is finally solved with good quality, I would easily believe, but not too much beyond this problem. Even this relatively easy problem requires very good qualification from developers.
—SA
Submit an article or tip