Real-World Noise-Robust Voice Interface
mpWAV's AI speech preprocessing reduces the impact of noise and echo without distorting the target voice, improving speech recognition for robots, kiosks, vehicles, and smart devices.
We improve the quality of the input signal first — without replacing your existing speech recognition engine.
Robots · Kiosks · Vehicles · Smart Devices · Factory Acoustic Inspection · Meetings · Hearing Assistance
Input: speech with noise and echo
→ mpWAV: mpAEC · mpBeamforming · mpAB · mpASR
→ Result: target voice preserved, noise reduced
→ Integration: applied directly to your existing ASR engine
→ Outcome: more accurate speech recognition
CES
CES 2024 Innovation Awards, two categories
Voice technology recognized in the global market
NET
NET New Technology Certification
Voice interface preprocessing for real environments
Procurement
Innovative Product for Public Procurement
A foundation for public sector deployment
Award
Prime Minister's Award, Korea Invention Patent Exhibition
Multi-microphone speech processing technology
Research
25+ years
Real-world voice interface preprocessing research
IP
Patented technology at home and abroad
Core speech processing IP and productization capability
Experience
Robot · kiosk · mobility PoCs
Validated in real products and real sites
Speech recognition has advanced rapidly, but real product environments contain several kinds of noise and echo at once.
Nearby conversation, TV and audio playback, road noise, machinery, room reverberation, and far-field speech all degrade the input quality delivered to the recognition engine.
Multiple voices arrive at the same time.
Sound played by the device loops back into the microphone.
Driving and wind noise interfere with voice commands.
The product's own operating sound mixes into the input.
The farther the user is from the microphone, the more noise dominates.
Reflections from walls and ceilings degrade recognition.
The problem is not only the recognition engine — it is the quality of the signal it receives.
Conversation | TV/Audio Echo | Road Noise | Machinery | Distance | Reverberation
When ordinary noise removal damages the user's voice along with the noise, recognition can actually get worse.
mpWAV focuses on reducing noise and echo while preserving the target voice recognition depends on.
Instead of relying on pre-registered microphone positions, mpWAV optimizes from the actual input signals.
That reduces the repeated tuning burden when microphone count or placement changes.
There is no need to replace the recognition engine you use today.
mpWAV preprocessing sits in front of it and improves the input signal quality.
Beyond algorithms, we support multi-channel audio I/O, DSP and FPGA porting, board design, and expansion to a dedicated SoC.
Ordinary: noise↓ + voice damaged | mpWAV: noise↓ + voice preserved
Multi-channel acoustic echo cancellation
Removes echo generated by the device itself — TV, car audio, voice prompts — that loops back into the microphone.
Multi-microphone ambient noise reduction
Uses multiple microphone inputs to reduce ambient noise and strengthen the target voice.
Integrated speech preprocessing
Integrates mpAEC and mpBeamforming to process speech in real time where noise and echo coexist.
On-device end-to-end speech recognition
A speech recognition solution that runs on servers, PCs, smart devices, and IoT/edge environments.
Microphone array & audio module design
We design the multi-channel audio I/O devices and microphone array modules that let the algorithms work in real products.
Mic Array HW → mpAEC → mpBeamforming (= mpAB) → mpASR → Product / Service
Helps service, home, care, and guide robots recognize user commands reliably amid ambient noise.
Get in touchVoice interfaces for barrier-free kiosks in stores, hospitals, transit, and public institutions.
Get in touchImproves voice command input where driving noise, wind, and car audio coexist.
Get in touchHelps TVs, appliances, smart home and IoT devices work reliably in real homes.
Get in touchAnalyzes acoustic signals to detect abnormal motor and equipment sounds on noisy production lines.
Get in touchVoice interfaces and dedicated systems for environments with severe ambient noise.
Get in touchImproves input quality for meeting minutes, source separation, game voice chat, and communications.
Get in touchClearSense Audio helps make everyday conversation clearer with ordinary earphones and a smartphone.
See ClearSense AudioRobot | Kiosk | Vehicle | Appliance | Factory | Defense | Meeting | Hearing
Voice technology performance varies with microphone configuration, user distance, noise type, and the recognition engine.
mpWAV validates before-and-after differences against real data and real product environments.
Before
Original speech with noise and echo
After mpWAV
Processed speech with the target voice delivered clearly
Demo audio is being prepared. Contact us to hear it today.
[▶ Listen: before] [▶ Listen: after]
Original input
“How's the weather today?”
Recognition of original
“How's the whether to…”
Recognition after mpWAV
“How's the weather today?”
0 dB input SNR · average of 100 utterances (target ≥ 10 dB)
0 dB input SNR · average of 100 utterances (target ≥ 15 dB)
Real-world 5 dB SNR · 65 dB speech, 60 dB noise · average of 100 utterances (target ≥ 25 dB)
Below 1 means real-time capable (target < 1)
Telecommunications Technology Association (TTA) test report TTA-25-1103. The article under test was our own speech enhancement solution, ClearSense Audio v1.1.0 (algorithm v1.0.0), tested 9–12 June 2025 and issued 17 November 2025. Corpora: LibriSpeech ASR and AIHub Korean Speech, with reverberation simulated using gpuRIR.
Robot voice interface preprocessing
Validated preprocessing to improve a home robot's speech recognition amid TV audio, conversation, and household noise.
Speech recognition in showroom noise
Evaluated a robot's voice interface in a showroom with crowd conversation and room reverberation.
Care robot voice interface
Applied speech recognition preprocessing to support voice interaction between older adults and a care robot.
Microphone-array kiosk module
Applied a multi-channel microphone I/O and preprocessing module to achieve high recognition performance in store noise.
Production line acoustic inspection
Validated detecting abnormal motor sounds in a noisy factory without a separate soundproof room.
Everyday conversation support in noise
Validated the smart listening app with users at the Bangbae and Nonhyeon senior welfare centers.
mpWAV's official material presents robots, kiosks, mobility, factory acoustic inspection, and ClearSense Audio pilots as its principal PoCs.
A module integrating the software with multi-channel microphone hardware.
Module inquiryWe port the technology to DSP, FPGA, or AP to meet your compute and latency requirements.
Porting inquiryLicense speech preprocessing or acoustic anomaly detection for your products and services.
Licensing inquiryDedicated chips and semiconductor IP collaboration for volume production and product expansion.
SoC partnership inquiryWe review the product, microphone configuration, user distance, dominant noise, and current errors.
We review your real speech data, or a jointly designed test environment.
We compare WER, command success rate, latency, and compute before and after preprocessing.
We apply the technology as software, a module, or a DSP/FPGA architecture.
We verify performance where the product is actually used and fine-tune.
We expand to follow-up models, product lines, regions, and languages.
Smart Listening App
ClearSense Audio is a smart listening app that refines the sound arriving at your smartphone with AI, helping you hear conversation in noise more clearly through ordinary earphones.
Its usability was reviewed in real environments through a Seoul Metropolitan Government assistive-technology program and pilots at senior welfare centers.
ClearSense Audio is a software service that assists everyday conversational listening. It does not provide medical diagnosis or treatment and does not replace hearing aids.
App screens + earphone usage scene
mpWAV solves real-world voice problems with speech processing research, AI software, multi-channel hardware design, and product integration capability.
Paper, patent, and project counts are the cumulative research record of the founding researchers.
Lab · FPGA/DSP boards · Mic array prototypes
Tell us your product type, microphone configuration, usage environment, and current errors — we will recommend the right validation method and integration structure.
You don't have to commit to a large project up front. Start a PoC with your most important environment and data.
mpWAV develops and licenses voice interface technology that makes speech clear in noisy, real-world environments, built on 25+ years of speech signal processing research.
(주)엠피웨이브·사업자등록번호 859-81-01905·대표 박형민
Teilhard Hall 405, 35 Baekbeom-ro, Mapo-gu, Seoul, Korea
© 2026 (주)엠피웨이브. All Rights Reserved.