Device and Method for Hostile Blind Bandwidth Extension of End-To-End Using One or More Convolutional Networks And/or Recurrent Network

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

This invention describes a device and method that take a narrow band (limited frequency) voice signal and use neural networks to convert it into a wide band (fuller, more natural-sounding) voice signal. Essentially, it makes low-quality voice audio sound richer and clearer by intelligently reconstructing missing frequency components via convolutional and/or recurrent neural networks.

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

• Improving voice call quality in telecommunication systems (e.g., mobile networks, VoIP) using legacy devices with limited bandwidth
• Enhancing archived audio recordings by restoring or expanding their quality
• Improving real-time voice transmission in walkie-talkies or emergency communication systems
• Upgrading speech recognition systems with improved input audio quality
• Restoring degraded voice signals in security and surveillance recordings

BenefitsContent extracted from patent full text and abstract with AI.

• Improves the intelligibility and naturalness of narrow band voice audio
• Enables better user experience in voice communication, even on bandwidth-limited networks or devices
• Utilizes AI-based neural networks for more accurate and adaptive enhancement compared to simple filtering techniques
• Easily integrable as a pre- or post-processing component in existing audio systems
• Potentially reduces the need for costly hardware upgrades by enhancing quality via software

Inventors & Applicants

Patent Abstract

To provide a device for processing a narrow band voice input signal and obtaining a wide band voice output signal by extending a band of the narrow band voice input signal.SOLUTION: A device for processing a narrow band voice input signal and obtaining a wide band voice output signal comprises: a signal envelope extrapolator (120) including a first neural network (125) for inputting a plurality of samples of a signal envelope of the narrow band voice input signal and outputting a plurality of extrapolated samples of the signal envelope; an excitation signal extrapolator (130) for inputting a plurality of samples of excitation signals of the narrow band voice input signal and outputting a plurality of extrapolated excitation signal samples; and a combiner (140) for generating the wide band voice output signal so that the wide band voice output signal extends a bandwidth for the narrow band voice input signal by depending on the plurality of extrapolated samples of the signal envelope and the plurality of extrapolated excitation signal samples.SELECTED DRAWING: Figure 1