Researchers demonstrate algorithm that has learnt how to predict sound
Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed an algorithm that has learnt how to predict sound. In the future, researchers predict that future versions of similar algorithms could be used to automatically produce sound effects for movies and TV shows.
Study’s lead author Andrew Owens said that such an algorithm that simulates sounds can unveil vital information about objects’ shapes and material types. In the study, the research team has used techniques from the field of ‘deep learning’.
As per the researchers, deep learning techniques are useful as they allow computers to go through massive amounts of data to find patterns on their won. The first step for an algorithm to produce sounds is to give it sounds to study.
The process continued for many months, the researchers recorded around 1,000 videos of an estimated 46,000 sounds representing different objects being hit and scrapped. In the second stage, the researchers fed those videos into a deep-learning algorithm.
“To then predict the sound of a new video, the algorithm looks at the sound properties of each frame of that video, and matches them to the most similar sounds in the database”, said Owens, who further affirmed that once the algorithm has the samples of audio, it stitches them to come up with a solid sound.
The researchers said that the study is a right step to mimic learning the way humans do, which is by integrating sound and sight. Another benefit of the work is that the team has come up with a library of 46,000 sounds, which is free and available for other researchers to use. As per the researchers, there is still scope to improve the system.
Frank Edelblut, a businessman who has no...Read More
Doubling down on President Donald Trump’s baseless...Read More
U. S. President Donald Trump reportedly insisted...Read More
A bill designed to make it easier for New Hampshire...Read More
New Hampshire legislators are mulling a package of...Read More
A team of detectives from Massachusetts State...Read More