AI can separate the musical instrument from the video just in

AI researchers at MIT's computer science lab have created AI that can separate musical instruments from videos with a simple click.

By using neuroscience and deep learning, researchers trained AI systems called PixelPlaye r with over 60 hours of continuous video. In those videos, AI is listening to musicians playing instruments.

The result, after about 60 hours of continuous training, is that AI can identify more than 20 different instruments even when they don't know what it is. As soon as the sound of the instrument is heard, the AI ​​can automatically isolate the instrument from the general sound of the video.

According to Gizmodo, listeners simply need to click on the instrument they want to hear in the video, AI will automatically process and detach that sound for a short time.

PixelPlayer uses deep learning to self-monitor, researchers at MIT are not sure how this AI learned to distinguish sounds from musical instruments.

In the future, AI-supported software is expected to revolutionize the music industry.

According to Zhao, the AI ​​system seems to recognize the true element of music, such as the sound frequency of an instrument. Zhao adds that if applied on robots, they can help robots easily capture their surroundings through sounds of vehicles and animals.

The team believes that if the AI ​​continues to be improved with the ability to identify differences between instruments in the same family, it will help remixes and instrumental performances become better. For example, the sound of the trumpet can be increased or the sound of piano can be reduced a little . All these random changes will create strange and extremely attractive performances.

Or with AI, musicians can easily focus on a specific part of the song, thereby easily handling the still entangled sections of the song. In the future, AI-assisted software is expected to revolutionize the music industry, improve remix techniques and create more interesting mixes.