A group of researchers from Carnegie Mellon University in Pittsburgh think they’ve figured out how to. They call their method a “brain-imaging approach” and claim it’s the largest evidence yet that the smartphone can be used to perform a full, non-trivial brain fingerprinting operation (a.k.a. brain-scanning.)
So who’s behind it? You can thank this person:
According to the researchers, the technique, called “vocal recognition,” will work with any song, and can use a single fingerprint to identify the song that is being sung.
The process involves using an application to identify which portion of the vocal sound is present in a recording (like a single chord, for example), and a special computer will then “scan” that portion of the track and then tell the app which instrument the artist is singing. Once that is done, the app will automatically spit back back a fingerprint that, the researchers claim, can be compared to those of other audio fingerprints taken from singers and recorded by other audio experts to be sure that you’re singing the song correctly.
(The researchers claim it’s “an area in the brain associated with facial recognition, facial expression, auditory processing (tone/frequency, rhythm, pitch), perception of sound, and emotion,” according to their research.)
The researchers say their “surgical approach” will be useful in detecting certain cancers and other genetic abnormalities from DNA analysis and other tests.
In other words, Google is using it. But it hasn’t announced any plans to release a Siri app yet.
You can read the researchers’ paper online, but below are the relevant key points, as described by WIRED:
Voice recognition is a complex cognitive skill involving the use of a speaker’s voice and spoken text, which are then compared against each other and compared against the entire corpus of recorded musical and song recordings. Voice recognition requires that the user be within a microphone’s dynamic range (about 20 feet or about 3 meters in space) to obtain reliable results, which also depends on the speaker’s voice. In addition, the human vocal tract changes its position frequently during recording and playback, making it difficult to use voice recognition technology to accurately compare a user’s voice against the entire musical and song corpus. Additionally, a large number of factors – such as speaker position and speaker volume – can influence the accuracy of voice recognition, making it difficult to obtain accurate results using voice recognition. In fact, if a person uses their voice
nle choppa type of beats to learn rapidly login google, nle choppa type of beats to learn rapid stemi, rock n learn dinosaur raptors with wings, how to learn rap god lyrics fast part rap, how to learn rap god fast part easybib