Conversa Voice Suite
Conversa is a comprehensive full-duplex, hands-free telephony speech processing suite providing total control of uplink and downlink audio. Conversa’s embedded multimicrophone, full-duplex voice processing delivers the ultimate in sound quality for speakerphone and headset applications.
A key feature of Conversa is Machine Learning Noise Reduction (MLNR) which dramatically improves call quality in noisy environments, effectively removing non-speech sounds like keyboard clicks, barking dogs, sirens, etc. from the talker’s audio.
Checkout Conversa’s MLNR performance in our video demonstration for speakerphone use case and gaming headset use case.
Whether you’re designing a speakerphone, headset, TWS ear set, wearable or automotive voice solution, you need full-duplex voice processing that will perform in all noise types.
Please contact us for more information.
ソフトウェア詳細
図
Features
- Complete uplink and downlink audio processing
- Super-wideband audio up to Fs=32kHz
- High-quality sample rate conversion
- Acoustic echo cancellation
- 55dB total acoustic echo cancellation (AEC) including nonlinear processing (NLP)
- Machine Learning based noise reduction
- Adaptive multi-mic beamforming leveraging machine learning
- Dynamic compression
- Limiter and squelch
- Sidetone
- Parametric EQs
対応製品
- i.MX-RT500: i.MX RT500 Arm® Cortex®-M33、DSP、GPUコア搭載クロスオーバーMCU
- i.MX-RT600: i.MX RT600 Arm® Cortex®-M33およびDSPコア搭載クロスオーバーMCU
- i.MX-RT1060: i.MX RT1060：Arm® Cortex®-M7搭載クロスオーバーMCU
- i.MX-RT1170: i.MX RT1170：Arm® Cortex®コア搭載1 GHzクロスオーバーMCU
- LPC55S6x: High Efficiency Arm® Cortex®-M33-Based Microcontroller Family
設計・リソース
関連ソフトウェア
クイック・リファレンス ソフトウェア・タイプ.
2 ソフトウェア・ファイル
-
AI/MLモデリング・ツール
Voice Intelligent Technology Speech to Intent Engine
-
アプリケーション例 ソフトウェア
Application Software Pack: Conversa Voice Calling
注: より快適にご利用いただくために、ソフトウェアのダウンロードはデスクトップで行うことを推奨します。