Voice Recognition Modules: Working Procedure and Applications

Advancement in technology has developed an intelligent man-machine interface technique that facilitates computers or machines or robots to be operated using the voice commands of a human without using any input systems, such as keyboards or mouse. This human-machine interface (HMI) can be achieved using voice recognition modules. In this article, we will discuss the voice recognition modules along with their working procedure and applications.

Voice Recognition Module

Voice recognition is a technique that facilitates a natural and convenient human-machine interface using the voice recognition module. It extracts and analyses voice features of human delivered to a machine or computer through the mic. Voice recognition technique is classified into many types based on different criteria such as the scope of the users, a number of words used for recognition, the naturalness of speaking. If the voice recognition level is more than 95%, then only the voice recognition is practically used.

Voice Recognition Module’s Working Principle

HM2007 is a single-chip CMOS voice recognition module. It is an on-chip analog front end large-scale integrated circuit with voice analysis, speech recognition, and voice recognition system control processes. HM2007 can be operated in two modes: manual mode and CPU control mode.

In the manual mode of operation, the voice recognition module HM2007 is used to build a simple recognition system by connecting components such as a keypad, SRAM of 8Kbyte memory, and other components. If power is on, then the HM2007 starts an initialization process and, if the WAIT pin is L, then the HM2007 checks external memory: 8Kbyte SRAM – whether it is perfect or not. But, if the WAIT is H, then the memory check process will be skipped by the HM2007. After this initialization process, the HM2007 moves into recognition mode. In the recognition mode, if the WAIT pin is H, then the HM2007 is ready to accept the voice and begins the recognition process. If the WAIT pin is L, then the HM2007 doesn’t accept the voice that is to be recognized. Then training new patterns by clearing trained patterns are the two operations performed using function keys: TRN and CLR. All the patterns in the HM2007 can be cleared by entering number key 99 and pressing CLR.

In the CPU control mode of operation, several functions such as RECOG, TRAIN, RESULT, UPLOAD, DOWNLOAD, RESET functions of the voice recognition module HM2007 are provided in this mode. Similar to a manual mode of operation, in this mode also power on, recognition, training, resulting, uploading, downloading, and reset operations are performed based on different criteria.

Applications of Voice Recognition Module

Voice Controlled Robotic Vehicle

A voice-controlled robotic vehicle with a long-distance-speech-recognition system is designed to be controlled by using a human-machine interface and voice commands for the remote operations of the robotic vehicle. An 8051 microcontroller is used along with a voice recognition module or speech recognition module for achieving the desired operation. The direction of the robotic vehicle’s movement can be controlled by using voice commands or push buttons. The voice commands are sent by the RF from the transmitting end towards the receiving end. Thus, the robotic vehicle moves in forward, backward, left, or right directions based on the commands received by the receiver.

This movement of the robotic vehicle in a specific direction can be controlled using two motors that are interfaced with the 8051 series microcontroller. RF transmitter converts the commands by a switch press or voice commands that are converted into encoded digital data for the benefit of the acceptable range (up to 200 meters) from the robotic vehicle. The encoded data received at the receiver circuit is decoded to being sent to another microcontroller for driving DC motors using motor driver IC for controlling the direction and movement of the motors using the decoded data of the voice commands.

This voice-controlled robotic vehicle with long-distance speech recognition projects can be integrated with a DTMF technology that facilitates the control of the robotic vehicle using a cell phone. This DTMF technology provides a very long-range communication compared to the RF technology – thus robotic vehicles can be controlled remotely from a very long distance.

Voice Recording and Playback Circuit

APR 9301 IC

APR 9301 IC consists of 28 pins and nonvolatile flash memory. It facilitates the recording of 100K cycles and memory storing for around 100 years. Only low voltage 5V and 25mA of current are required for the normal working operation of the APR 9301 integrated circuit.

Working of the Voice Recording and Playback Circuit

APR 9301 IC performs voice recording and playback operations. Recording operation can be performed by receiving voice signals using good quality (of any kind) condenser mic connected to the 17 and 18 pins of an integrated circuit. If we close the switch S1, then the recording mode can be started for easily recording a voice message for 20-30 seconds. As an indication of recording LED connected to pin 25 of APR 9301 IC will blink as long as L pin 27 is grounded.

After the completion of 20 cycles with the last memory, the recording process terminates automatically by changing the value of a resistor R1 connected to the pins 6 and 7 of the APR 9301 IC. A maximum recording time durations of 20 seconds, 24 seconds, and 30 seconds can be obtained by changing the values of the resistor R1 as 52K, 67K, and 89K, respectively.

The input section is muted automatically during the playback mode. If the switch S2 is closed, then a message will come from the speaker from the beginning of the recorded messages. If recording or playback functions are completed, then the APR 9301 IC will enter into a standby mode.

This circuit can be accumulated on a common PCB. Solder the APR 9301 IC base carefully such that no shorting should occur between the pins of the integrated circuit. Check the circuit after assembling it on the PCB, then connect the IC in the IC base. Before giving power supply to the circuit, closely check the pin connections. 5-volt regulator IC-based Power supply is used for giving power supply to the circuit. A 2-inch-8-ohms good quality speaker is used to get a clear sound. Recording can be done by pressing the switch S1. The sound signals (Speech or Music) are picked up by the mic and are passed to the IC wherein the voice signals are stored in the memory cells. If we close the switch S2, then the playback starts and we can hear the recorded message through the speaker.

Voice recognition module can be used in many applications such as for controlling the aircraft systems using voice commands of the pilots, for controlling motorized wheel car using voice-activated multiprocessor, and so on. If you are interested to know about the electrical and electronics projects based on the voice recognition module, then you can approach us by posting your comments in the comment section below.

Photo Credits:

• APR 9301 IC electroschematics
• HM 2007 the4cs
• voice recognition module emartee