18605688. Spatial Audio Upscaling Using Machine Learning simplified abstract (Apple Inc.)
Contents
Spatial Audio Upscaling Using Machine Learning
Organization Name
Inventor(s)
Ismael H. Nawfal of Redondo Beach CA (US)
Symeon Delikaris Manias of Los Angeles CA (US)
Mehrez Souden of Los Angeles CA (US)
Joshua D. Atkins of Lexington MA (US)
Spatial Audio Upscaling Using Machine Learning - A simplified explanation of the abstract
This abstract first appeared for US patent application 18605688 titled 'Spatial Audio Upscaling Using Machine Learning
The patent application describes a method for converting first order Ambisonics (FOA) audio into a desired higher order Ambisonics (HOA) format using machine learning.
- Processor formats FOA audio signals into a stream of audio frames.
- Machine learning model reformats the FOA audio into the desired HOA format.
- Output audio in the desired HOA format can be rendered for playback in a chosen format.
Potential Applications: - Virtual reality and augmented reality audio experiences - Immersive audio production for film, television, and gaming - Spatial audio processing for live events and concerts
Problems Solved: - Simplifies the conversion process from FOA to HOA audio formats - Enhances the quality and realism of spatial audio rendering
Benefits: - Improved spatial audio accuracy and immersion - Streamlined workflow for audio professionals - Enhanced user experience in virtual environments
Commercial Applications: Title: Spatial Audio Conversion Technology for Immersive Experiences This technology can be utilized in VR/AR content creation, audio post-production studios, and live event sound engineering to enhance spatial audio quality and realism.
Questions about the technology: 1. How does this technology improve the spatial audio experience for users?
- The technology enhances the accuracy and immersion of spatial audio by converting FOA audio into a desired HOA format.
2. What are the potential applications of this technology beyond entertainment?
- This technology can also be used in acoustic simulations, architectural design, and teleconferencing to create immersive audio environments.
Original Abstract Submitted
A sound scene is represented as first order Ambisonics (FOA) audio. A processor formats each signal of the FOA audio to a stream of audio frames, provides the formatted FOA audio to a machine learning model that reformats the formatted FOA audio in a target or desired higher order Ambisonics (HOA) format, and obtains output audio of the sound scene in the desired HOA format from the machine learning model. The output audio in the desired HOA format may then be rendered according to a playback audio format of choice. Other aspects are also described and claimed.
