Xilicom Technology

Technology: Core IP Licensing

Application: Adaptive Voice Communication

Welcome to our Technology Research in San Jose, California.

Pictures are worth a thousand words, but only sounds tell a story.

We have developed a family of adaptive audio processing algorithms that selectively enhance sounds of interest while suppressing unwanted noise. The algorithm heuristically learns familiar voices and picks them out from complex, noisy environments — even in the presence of machinery, wind, and crowd noise.

One of the most compelling applications of this technology is occupational safety. Applied to industrial earmuffs, the algorithm minimizes harmful noise exposure while allowing workers to hear conversation, alarms, and situational cues clearly and naturally.

Listed below are some of its features and many possible applications. Please let us know if you are interested in licensing these technologies.

Jason Kim
4/7/2010

Application: Adaptive Audio Amplification

Selectively amplifies desired sounds — voices, alerts, ambient cues — while attenuating harmful noise levels in real time.

Continuous real-time noise floor estimation
Adaptive gain control per frequency band
Low-latency processing for natural sound perception

Application: Augmented Hearing Protection

Industrial-grade hearing protection that does not sacrifice situational awareness. Workers remain protected from damaging sound levels while maintaining full auditory communication with their team.

Meets NIOSH and OSHA noise exposure guidelines
Compatible with 2G/3G/4G cellular communication
Earmuff and in-ear form factors

Technology 1: Single Mic Noise Reduction

Single Mic:

Analyze,

Separate,

Enhance,

Our single-microphone noise reduction algorithm uses spectral subtraction and statistical signal modeling to separate speech from background noise. It analyzes the incoming audio signal in real time, estimates the noise floor during non-speech intervals, and applies adaptive filtering to enhance voice clarity.

This approach works well in environments with relatively stationary noise sources, such as HVAC systems, engine hum, or factory floor machinery.

Technology 2: Multiple Mic Noise Reduction

Multi Mic:

Locate,

Beamform,

Isolate,

Our multi-microphone noise reduction system uses spatial signal processing and beamforming to localize and isolate desired sound sources. By analyzing the phase and amplitude relationships between microphone signals, it steers a virtual acoustic beam toward the speaker of interest and nulls interfering sources.

This approach excels in highly dynamic or non-stationary noise environments such as construction sites, factory floors with multiple machines, or outdoor jobsites with wind and crowd noise.