GANStrument

Introducing GANStrument: A New Way to Create Sounds

GANStrument is an innovative sound generation application developed by Sony CSL, utilizing the power of cutting-edge AI models to create captivating new sounds inspired by existing ones. With GANStrument, you can effortlessly blend the characteristics of two input sounds to generate unique soundscapes, offering an unparalleled and intuitive sound creation experience that surpasses traditional synthesizers.

Blend and Create: Crafting Your Perfect Sound

For example, if you desire a sound that combines the rich tones of an organ with the attack of a guitar, simply input the organ and guitar sounds into GANStrument and blend them together. This creates a sound that embodies the best of both instruments. Users can control the blending ratio within the application, allowing them to craft the precise sound they envision. This leads to a more intuitive sound creation process compared to conventional synthesizers.

Expanding Creative Possibilities: Beyond Musical Instruments

Furthermore, GANStrument can incorporate not only musical instrument sounds but also natural and pitchless noises, broadening creative possibilities even further. For instance, if you wish to add background music to a video featuring a cat, GANStrument enables you to generate a new instrument sound that includes elements of the cat’s meow. By adding contextual elements to the sound, GANStrument enhances the range of musical expression and sound creation.

Experience the Future of Audio Innovation with GANStrument

GANStrument is a captivating application that harnesses Sony CSL’s state-of-the-art AI models, providing users with a unique and engaging sound creation experience. If you’re interested in exploring the creative possibilities that GANStrument has to offer, don’t hesitate to get in touch. For more information on GANStrument’s AI model and the research behind it, you can view the research paper through the following link.

Related Publication

ICASSP 2023
GANStrument: Adversarial Instrument Sound Synthesis with Pitch-Invariant Instance Conditioning
Gaku Narita, Junichi Shimizu, Taketo Akama
neural synthesizer, generative adversarial networks, adversarial feature extraction

Patent
Inventors: Gaku Narita, Junichi Shimizu, Taketo Akama, Shintaro Oguchi, Kohei Yamamoto, Haruhiko Kishi
Status: Application Pending