Voice Recognition Solution
Voice recognition solution with high recognition rate even in noisy environment
The voice recognition is used as a human machine interface which has been adopted in lots of products like robot and smart speaker. The voice recognition is developed from the needs to adopt more convenient functions while keeping costs as low as possible in consumer equipment and industrial equipment. And the voice recognition function is becoming an important additional feature because it is possible to help visually impaired and elderly persons by using the voice recognition function. Renesas is providing a voice recognition solution which does not need internet connection (Edge voice recognition solution) to make current products differentiation and high functionality.
Implemented by A/D converter or SSI (Serial Sound Interface) and middleware
Realize high recognition rate under noisy environment using “Noise suppressor technology”
(ex) noise suppressor technologies
- Beamforming
Reducing noise from other of target - Noise suppressor
Reducing steady noise - Echo cancellation
Preventing or removing echo that is being created or already present
Simple Voice Recognition Solution
Edge voice recognition solution using noise suppressor technologies
■ Features
- Small Voice recognition solution with MEMS microphone
- LED ON/OFF and controls Infrared communication compatible devices by Infrared communication(*) according to the recognition result
- Easily change Voice recognition parameters by checking the voice waveform with the evaluation tool
* Available for RX231 Voice Recognition solution
| RX231 Voice Recognition Solution | RX651 Voice Recognition Solution | RA6M1 Voice Recognition Solution | ||
|---|---|---|---|---|
|
|
|
||
| Hardware | MCU | RX231 (R5F52318ADFL) ROM/RAM：512KB/64KB Package：48pin LQFP |
RX651 (R5F5651EDDFM) ROM/RAM：2MB/640KB Package： 64pin LFQFP |
RA6M1 (R7FA6M1AD3CFM) ROM/RAM：512KB/256KB Package： 64pin LQFP |
| Microphone | Digital MEMS Mic x2 | Analog MEMS Mic x2 | Analog MEMS Mic x2 | |
| Other functions | Infrared communication, RGB LED, USB(Full Speed), Push switch | RGB LED, USB(Full Speed), Push switch | RGB LED, USB(Full Speed), Push switch | |
| Board size | 60mm x 40mm | 60mm x 40mm | 60mm x 40mm | |
| Software | OS | Not used | Not used | Not used |
| Middleware | Advanced Media/AmiVoice Micro Techno Mathematical/Zoom Voice |
Advanced Media/AmiVoice Micro Techno Mathematical/Zoom Voice |
Advanced Media/AmiVoice Micro Techno Mathematical/Zoom Voice |
|
| - | - | Toshiba Digital Solutions/RECAIUS™ Voice Trigger Techno Mathematical/Zoom Voice |
||
■ Reference designs
| Hardware | Software (Source code & Application notes) |
Voice recognition evaluation tool | |
|---|---|---|---|
| RX231 Voice Recognition Solution | RX231 Group Voice Recognition Demo Board R12AN0096EJ0101 | Contact to Renesas sales office for the detail information | Contact to Renesas sales office for the detail information |
| RX651 Voice Recognition Solution | RX651 Group Voice Recognition Demo Board R12AN0104EJ0101 | Contact to Renesas sales office for the detail information | Contact to Renesas sales office for the detail information |
| RA6M1 Voice Recognition Solution | RA6M1 Group Voice Recognition Demo Board R12AN0103EJ0101 | Contact to Renesas sales office for the detail information | Contact to Renesas sales office for the detail information |
RA6M3 HMI solution
A solution that realizes Edge voice recognition, Voice playback, Capacitive Touch operation and Environmental sensing with “RA6M3 MCU” 1chip
■ Features
- Realize Voice recognition, Voice playback, TFT LCD control and Environmental sensor control by RA6M3 1chip
- Change the TFT LCD and Voice feedbacks according to the recognition result
- Easily change M/W parameters while checking the voice waveform with the evaluation tool
| RA6M3 HMI solution | ||
|---|---|---|
|
||
| Hardware | EK-RA6M3G
|
・MCU：RA6M3 (R7FA6M3AH3CFC) ・USB (Debug, Full Speed, High Speed) ・Graphics Expansion Board |
| HMI Expansion Board
|
・Analog MEMS Mic x2 ・External expansion microphone circuit (MEMS type (Analog output) Or Electret condenser type) ・Speaker operation circuit & Speaker ・Humidity and Temperature Sensor(RENESAS/HS3001) ・Gas Sensor(RENESAS/ZMOD4410) ・Arduino Uno Connection |
|
| Software | OS | Amazon Free RTOS |
| Middleware | Advanced Media/AmiVoice Micro Techno Mathematical/Zoom Voice CRI Middleware/D-Amp Driver |
|
| Toshiba Digital Solutions/Voice Trigger Techno Mathematical/Zoom Voice CRI Middleware/D-Amp Driver * The voice playback file was created by Toshiba Digital Solutions/RECAIUS speech synthesis middleware Text-to-Speech |
||
■ Reference designs
| Hardware | Software (Source code & Application notes) |
Voice recognition evaluation tool | |
|---|---|---|---|
| RA6M3 HMI solution | RA6M3 Group RA6M3 HMI Expansion Board R12AN0106EJ0100 |
Contact to Renesas sales office for the detail information | Contact to Renesas sales office for the detail information |
High performance HMI solution
Realized voice recognition, voice synthesis and touch panel by using 1-chip RZ/A1H without internet connection.
■ Features
- Voice recognition solution with noise suppressor technology
- Realize Voice recognition, Speech synthesis and TFT LCD control by RZ/A1H 1chip
- Feedback the results by voice synthesis function and LCD display function
Click here for details of RZ/A1H
| Function | Partner | Middleware |
|---|---|---|
| Noise suppressor | Techno Mathematical Co., Ltd. | Zoom Voice |
| Voice recognition | Advanced Media Inc. | AmiVoice Micro |
| Voice synthesis | Hitachi ULSI Systems Co., Ltd. | Ruby Talk® |
■ Features
Enabling below function by connecting evaluation board to the PC
- Visually check the sound input with a waveform
- Change the M/W parameters for voice recognition and noise reduction
- Display recognized ID
- Sound data before and after noise processing can be saved and played back
Voice Recognition Middleware：
Advanced Media/AmiVoice Micro
Features
- Realized voice recognition in none internet connection, low clock and small memory environment compared to existing products
- Two acoustic models
- Normal model
- High recognition model
- Support of VAD (voice activity detection)
- It includes a module that detects sections of only human speech from any voice, and the detection sensitivity can be adjusted according to usage scenes and tasks.
- RXv2 CPU-based RX family (RX231, RX230, RX65N, RX651, RX64M Group, etc.)
- RXv3 CPU-based RX family (RX72M, RX72N Group, etc.)
- ARM Cortex-M4(RA6M1, RA6M2, RA6M3 Group, etc.)
- ARM Cortex-A9(RZ/A1H, A1L Group, etc.)
Supported MCU
Required memory size
- Normal model
ROM: over 33[KB], RAM: over 23[KB] - High recognition model
ROM: over 482[KB], RAM: over 23[KB]
Required ROM/RAM size against to recognition word number
| Number of Words | Normal model [KB] | High recognition model [KB] | ||
|---|---|---|---|---|
| ROM | RAM | ROM | RAM | |
| 5 | 33 | 23 | 482 | 23 |
| 10 | 54 | 25 | 681 | 25 |
| 20 | 78 | 28 | 995 | 28 |
| 30 | 96 | 30 | 1,226 | 30 |
| 40 | 109 | 33 | 1,444 | 33 |
| 50 | 117 | 33 | 1,587 | 33 |
| 100 | 143 | 46 | 2,143 | 46 |
| 150 | 160 | 55 | 2,452 | 55 |
* Information for reference (It changes according to the language and the content of recognition word.)
Languages
- Normal model
Japanese, English, Chinese (Mandarin), Thai - High recognition model
Japanese
Voice Recognition Middleware ：
Toshiba Digital Solutions/RECAIUS™ Voice Trigger
- RECAIUS Voice Trigger realize voice control function without internet connection.
- User can change target phrases without speech data and use this as a customized detector of your own wake-words and/or voice commands.
- ARM Cortex-M4 (RA6M1, RA6M2, RA6M3 Group, etc.)
- ARM Cortex-A9 and later
- * Information for reference (It might be changed according to the language and its words.)
- Japanese, American English and Mandarin Chinese
- To be commercialized (Available for evaluation):
Canadian French, American Spanish, British English, French, German, Spanish, Italian
Features
Supported MCU
Required Memory Size
| Number of Words | ROM [KB] | RAM [KB] |
|---|---|---|
| 5 | 145 | 45 |
| 10 | 160 | 50 |
| 20 | 190 | 65 |
Support Language
Noise suppressor Middleware：
Techno Mathematical /Zoom Voice
Features
Support two noise suppressor technologies
Beam forming
- Extracting the target sound properly from front with reducing the background noise
- Using two non-directive microphones
- Effect could be set from “1: weak to 7: strong”
Noise suppressor
- Noise reduction 30dB (about 1/30) max.
- Noise reduction could be set according to frequency
High speed process version applied DSP instruction
The processing speed of DSP instruction applied version is 30% higher
Supported MCU
DSP instruction applied version：
RXv2 CPU-based RX family (RX231, RX230, RX65N, RX651, RX64M Group, etc.)
RXv3 CPU-based RX family (RX72M, RX72N Group, etc.)
Normal version：
ARM Cortex-M4 (RA6M1, RA6M2, RA6M3 Group, etc.)
ARM Cortex-A9 (RZ/A1H, A1L Group, etc.)
Required memory size
ROM: 40[KB], RAM: 10[KB]
Use case of beam forming and noise suppressor
The high recognition rate is achieved even under noisy environments by using Zoom Voice.
Especially very high effect can be expected at 5[dB] or less S/N ratio.
The recognition rate by using Zoom Voice under noisy environments (AmiVoice Micro is used for voice recognition)
Note 1, using sound of vacuum cleaner and washing machine as the source of noise.
Note 2, this data is base on the research of Renesas.
Development and sales of voice recognition software products
Advanced Media. Inc
System Integration, Development, Manufacture and Sales of ICT Solutions Utilizing IoT and AI Technology
Toshiba Digital Solutions
CONTACT：https://www.toshiba-sol.co.jp/en/contact/index.html
Email: tdsl-recaius-mw-sales-r1@ml.toshiba.co.jp
Development and sales of image, acoustic and sound processing software and hardware products