Signal separator, method for determining output signals on the basis of microphone signals, and computer program

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention concerns a signal separator system—alongside corresponding methods and computer programs—that can distinguish and extract specific audio sources (like a desired voice or sound) from multiple microphone recordings, even in the presence of interfering background sources. Unlike traditional blind source separation, which sometimes loses important spatial cues or can't fully separate interference, this system limits signal distortion during extraction and maintains correct spatial information, ensuring the separated output preserves natural positioning and clarity of both the primary and residual sources.

Use CasesContent extracted from patent full text and abstract with AI.

• Enhancing voice clarity in conference or meeting rooms by removing background noise while preserving spatial cues.
• Improving audio quality in hearing aids or binaural headsets, allowing users to focus on useful sounds (like speech) while suppressing interfering noises.
• Audio enhancement for telecommunication devices (smartphones, video calls), so users can be clearly heard in noisy environments.
• Source separation for professional audio recording or live sound engineering, isolating individual instruments or voices.
• Robotics and smart home systems that require accurate recognition and localization of sounds for voice-command interfaces or security monitoring.

BenefitsContent extracted from patent full text and abstract with AI.

• Effectively separates desired audio sources from interfering signals in multi-microphone setups with minimal signal distortion, preserving audio quality.
• Maintains the natural spatial characteristics of both the main sound and any residual background, aiding in clear spatial perception.
• Adaptive to different environments and does not require pre-known spatial locations of sound sources.
• Can be implemented in software (via computer program), making it suitable for integration into devices or real-time applications.
• Improves user experience in challenging acoustic environments by enhancing intelligibility and reducing listener fatigue.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

A signal separator, a method and computer product for determining a first output signal describing an audio content of a useful-signal source in a first microphone signal, and for determining a second output signal describing an audio content of the useful-signal source in a second microphone signal.