RPI ID: 2020-011-401

Innovation Summary:
This invention presents a method for separating multiple overlapping speech signals using a 3-tuple coprime microphone array. The system processes subarray data to isolate individual talkers in noisy environments. It enhances speech recognition and audio clarity in real-time applications. The approach is computationally efficient and suitable for embedded systems.

Challenges / Opportunities:
Separating overlapping voices in real-world environments is a major challenge for voice-controlled systems and hearing aids. Traditional beamforming techniques are limited in resolution and scalability. This invention offers a novel array configuration and processing method that improves separation performance. It opens opportunities in smart devices, assistive technology, and surveillance.

Key Benefits / Advantages:
✔ High-resolution separation distinguishes multiple speakers clearly
✔ Compact array design reduces hardware complexity
✔ Real-time capability suitable for live audio processing
✔ Versatile integration with consumer electronics and industrial systems

Applications:
• Smart speakers and voice assistants
• Hearing aids and assistive listening devices
• Surveillance and security systems
• Conference and telepresence systems

Keywords:
#Multitalkerseparation #coprimemicrophonearray #speechenhancement #audiosignalprocessing

Intellectual Property:
US Issued Patent 11,937,056 B2