TWI904664B - System and method for voice service with customer identification - Google Patents

Abstract

A voice service system with customer identification and method thereof are provided. The voice service system includes a server and a voice service device, the voice service device includes a processor, storage medium, a transceiver, a first side sound receiving and reproducing device, and a second side sound receiving and reproducing device. The processor is configured to transmit identity information of a customer to the server, to obtain the voice setting corresponding to the customer from the server; adjust at least one of the frequency value and the volume value of the voice setting, in response to recognizing semantic keywords in the voice signal received from the first side sound receiving and reproducing device; adjust the frequency and volume of the voice signal according to the voice setting; and transmit the voice signal to the second side sound receiving and reproducing device.

Description

Voice service system and methods with customer identification

This invention relates to a voice service technology, and more particularly to a voice service system and method with customer identification.

As Taiwan gradually moves towards a super-aged society, the service industry will face an increasing number of elderly customers. In addition to aging, the elderly may experience hearing loss at specific frequencies due to illness or other factors. Furthermore, the relatively noisy environment of service industry counters makes it difficult for them to communicate with staff when conducting business, resulting in a poor service experience.

In view of this, the present invention provides a voice service system and method with customer identification, which can adjust the frequency and volume of the service personnel's voice based on the customer's response to the service personnel's voice, so as to enable the customer and the service personnel to communicate smoothly, thereby improving the customer's service experience and saving the time required to handle business.

This invention discloses a voice service system with customer identification, comprising a server and a voice service device. The voice service device includes a processor, storage media, a transceiver, a first-side receiver/amplifier, and a second-side receiver/amplifier. The server stores voice settings. The transceiver is communicatively connected to the server. A processor coupled to a transceiver, a storage medium, a first-side amplification device, and a second-side amplification device, wherein the processor is configured to transmit customer authentication information to a server to obtain voice settings corresponding to the customer from the server; in response to identifying semantic keywords in a voice signal received from the first-side amplification device, adjusting at least one of the frequency and volume values of the voice settings; adjusting the frequency and volume of the voice signal according to the voice settings; and transmitting the voice signal to the second-side amplification device.

In one embodiment of the invention, the processor responds to the failure to identify semantic keywords from the speech signal without adjusting the speech settings.

In one embodiment of the present invention, the processor responds to receiving a speech signal from the first-side receiver/amplifier, determines whether a semantic keyword is identified from the speech signal, and, in response to identifying the semantic keyword from the speech signal, adjusts at least one of the frequency value and volume value of the speech setting.

In one embodiment of the invention, the processor responds to determining that the voice settings have been adjusted and sends the adjusted voice settings to the server.

In one embodiment of the present invention, the server stores a machine learning model, and the processor responds by identifying semantic keywords from the speech signal, transmitting the semantic keywords and speech settings as inputs to the machine learning model to the server to obtain suggested values for the corresponding speech settings; and adjusting at least one of the frequency and volume values of the speech settings according to the suggested values.

In one embodiment of the present invention, the processor associates all identified semantic keywords with the voice setting adjustment process and sends them to the server as training data, wherein the server trains or updates the machine learning model based on the training data.

In one embodiment of the present invention, when the processor sends the customer's identity verification information to the server and the server notifies that there is no corresponding voice setting for the customer, the processor sets the frequency and volume values obtained by analyzing the voice signal as the frequency and volume values of the voice setting.

In one embodiment of the present invention, the processor performs noise reduction on the speech signal and transmits it to the second-side amplification device, and performs noise reduction on the speech signal received from the second-side amplification device and transmits it to the first-side amplification device.

In one embodiment of the present invention, the above-mentioned voice service system further includes an ATM, and when a customer operates the ATM, the server performs an identity verification procedure on the customer through the ATM, and in response to the successful identity verification procedure, transmits the corresponding voice settings of the customer to the ATM.

The present invention provides a voice service method with customer identification, comprising the following steps: transmitting customer identity verification information to a server to obtain the corresponding customer's voice settings from the server; identifying semantic keywords in response to a voice signal received from a first-side receiver/amplifier and adjusting at least one of the frequency and volume values of the voice settings; adjusting the frequency and volume of the voice signal according to the voice settings; and transmitting the voice signal to a second-side receiver/amplifier.

Based on the above, the customer-identified voice service system and method of this invention can adjust the frequency and volume of the service personnel's voice based on the customer's response to the service personnel's voice, enabling smooth communication between the customer and the service personnel, thereby improving the customer's service experience and saving the time required to complete the transaction. On the other hand, for customers who have previously completed transactions, suitable voice settings can be recorded, so that the frequency and volume of the service personnel can be adjusted more efficiently when the customer completes transactions again, further enhancing the customer's service experience.

Figure 1 is a schematic diagram of a voice service system 10 with customer identification according to an embodiment of the present invention. In this embodiment, the voice service system 10 may include a voice service device 100 and a server 200. The voice service device 100 may include a processor 110, a storage medium 120, a transceiver 130, a first-side amplification device 141, and a second-side amplification device 142, wherein the processor 110 may be coupled to the storage medium 120, the transceiver 130, the first-side amplification device 141, and the second-side amplification device 142. Server 200 may include processor 210, storage medium 220, and transceiver 230, wherein processor 210 may be coupled to storage medium 220 and transceiver 230.

Processor 110 or 210 may be, for example, a central processing unit (CPU), a digital signal processor (DSP), other programmable general-purpose or special-purpose microprocessors, programmable controllers, application-specific integrated circuits (ASICs), or other similar components or combinations thereof.

Storage media 120 or 220 are respectively used to store various software, data, and code required for operation by processor 110 or 210. Storage media 120 is, for example, any type of fixed or removable random access memory, read-only memory, flash memory, hard disk, solid-state drive, or other similar components or combinations thereof, and the invention is not limited thereto. In this embodiment, storage media 120 may, for example, be used to store voice signals received from the first-side receiver 141 and the second-side receiver 142, as well as other required software, data, and code. Storage media 220 can, for example, store multiple voice settings containing corresponding customer voice settings, a machine learning model for outputting suggested values for adjusting corresponding customer voice settings, and other necessary software, data, and various types of code.

Transceiver 130 or 230 transmits and receives signals wirelessly or via wire, and can also perform operations such as low-noise amplification, impedance matching, mixing, up or down frequency conversion, filtering, amplification, and similar functions.

The first-side receiver 141 and the second-side receiver 142 may be, for example, a combination of a microphone and a speaker, or a combination of various audio receiving devices and various audio playback devices. In this embodiment, the first-side receiver 141 is used to receive the voice signals of the service personnel and play the voice signals of the customers, and the second-side receiver 142 is used to receive the voice signals of the customers and play the voice signals of the service personnel.

Figure 2 is a flowchart illustrating a method for applying a voice service system 10 with customer identification, according to an embodiment of the present invention.

In step S210, the processor 110 of the voice service device 100 of the voice service system 10 with customer identification transmits the customer's identity verification information to the server 200 to obtain the corresponding voice settings from the server 200. For example, the corresponding voice settings may include frequency and volume values that allow the customer to hear the spoken content clearly.

In step S220, the processor 110 responds to receiving a speech signal from the first-side amplification device 141 and determines whether semantic keywords are identified from the speech signal. Specifically, the storage medium 120 of the speech service device 100 may, for example, store a natural language processing program or language model capable of recognizing the utterance content corresponding to the speech signal. The processor 110 can obtain semantic keywords representing the speech signal that need to be adjusted by inputting the speech signal into the natural language processing program or language model. For example, a customer may experience hearing loss due to aging or illness, leading to complaints that they cannot clearly hear the service personnel or that the staff's voice is too soft during transactions. In response, the service personnel can say, for example, "The customer says they cannot hear clearly," "The customer is speaking too softly," or other similar semantic keywords indicating a need to adjust the voice signal. If the processor 110 determines that it has identified semantic keywords in the voice signal indicating a need for adjustment, it can execute step S230.

If, in step S220, the processor 110 determines that no semantic keywords have been identified from the voice signal, meaning that the voice signal adjusted by the current voice setting frequency and volume values is sufficient for the customer to clearly hear what the service personnel are saying, then the voice settings will not be adjusted. In this case, step S250 can be executed.

In step S230, processor 110 sends semantic keywords and voice settings as inputs to the machine learning model to server 200 to obtain suggested values for the corresponding voice settings. Specifically, server 200 stores a machine learning model that may, for example, receive the current frequency and volume values of semantic keywords and voice settings as inputs, and output suggested values including frequency and volume suggestions, which can be used to appropriately adjust the voice settings. The machine learning model may be, for example, a neural network model based on Generative Pre-trained Transformer 3 (GPT-3) or Long Short-Term Memory (LSTM), but is not limited to these.

In step S240, processor 110 adjusts at least one of the frequency and volume values of the voice setting based on suggested values. For example, when the suggested frequency value differs from the currently set frequency value, processor 110 adjusts the frequency value of the voice setting based on the suggested frequency value. When the suggested volume value differs from the currently set volume value, processor 110 adjusts the volume value of the voice setting based on the suggested volume value.

In step S250, processor 110 adjusts the frequency and volume of the voice signal according to the voice settings. Specifically, processor 110 adjusts the frequency and volume of the service personnel's voice signal received from the first-side receiver/amplifier 141 according to the voice settings. This makes it easier for customers to hear what the service personnel are saying.

In one embodiment, when processor 110 sends customer authentication information to server 200 and server 200 notifies that there is no corresponding voice setting for the customer, the frequency and volume values obtained by analyzing the voice signal received from the first-side receiver/amplifier 141 are set as the frequency and volume values of the voice setting. This allows for the matching of the frequency and volume values for customers making their first transaction.

In one embodiment, if the processor 110 determines that the frequency and volume values set for the voice signal are similar to those obtained by analyzing the voice signal received from the first-side receiver/amplifier 141 from the service personnel, it will not adjust the frequency and volume of the voice signal. For example, if the customer has no hearing loss, the adjustment of the frequency and volume of the voice signal can be omitted.

In step S260, processor 110 performs noise reduction on the voice signal. Processor 110 may, for example, employ active noise cancellation (ANC) technology to reduce noise in the voice signal. Specifically, processor 110 may reduce the noise in the voice signal of a service personnel received from the first-side receiver 141 before transmitting it to the second-side receiver 142, and may reduce the noise in the voice signal of a customer received from the second-side receiver 142 before transmitting it to the first-side receiver 141. This prevents noise near the counter from affecting customer business.

In step S270, processor 110 transmits the voice signal to the second-side receiver 142. Overall, the voice service device 100 adjusts the frequency and volume of the service personnel's voice signal received from the first-side receiver 141 and performs noise reduction processing so that the customer can clearly hear the content of the voice signal from the second-side receiver 142.

In step S280, the processor 110 responds to the determination that the voice settings have been adjusted and sends the adjusted voice settings to the server 200. For example, the adjusted voice settings include frequency and volume values suitable for the customer. When the customer conducts business again, the processor 110 can obtain the frequency and volume values suitable for the customer's voice settings from the server 200 to adjust the voice signal received from the first-side receiver/amplifier 141. In this way, the voice signal can be adjusted more efficiently.

In one embodiment, the voice service system 10 further includes an ATM, and when a customer operates the ATM, the server 200 performs an identity verification procedure on the customer through the ATM, and in response to successful identity verification, transmits the corresponding voice settings to the ATM. This allows the customer to clearly hear the content of the voice signal played from the ATM.

In one embodiment, the first-side amplification device 141 may be, for example, a telephone device on the service personnel's end. In this way, when a customer makes a customer service call, they can clearly hear the content of the voice signal received from the first-side amplification device 141.

In step S290, processor 110 associates all identified semantic keywords with the voice setting adjustment process and sends this as training data to server 200. In this way, server 200 can train or update the machine learning model based on training data containing actual customer feedback, thus ensuring that the suggested values output by the machine learning model better match the customer's feedback and avoiding outputting suggested values that the customer has already indicated were unclear.

In one embodiment, the processor 110 stores all voice signals received from the first-side receiver 141 and the second-side receiver 142 in the storage medium 120 and transmits them to the server 200 as a record and proof of business transactions between service personnel and customers, in order to reduce potential disputes.

Figure 3 is a flowchart illustrating a method for applying a voice service system 10 with customer identification, according to an embodiment of the present invention.

In step S310, the customer's identity verification information is transmitted to server 200 to obtain the corresponding customer's voice settings from server 200. In step S320, in response to the voice signal received from the first-side receiver 141, semantic keywords are identified, and at least one of the frequency and volume values of the voice settings is adjusted. In step S330, the frequency and volume of the voice signal are adjusted according to the voice settings. In step S340, the voice signal is transmitted to the second-side receiver 142.

In summary, the customer-identification voice service system and method of this invention can adjust the frequency and volume of the service personnel's voice based on the customer's response to the service personnel's voice, enabling smooth communication between the customer and the service personnel, thereby improving the customer's service experience and saving time spent on handling business. On the other hand, for customers who have previously conducted business, suitable voice settings can be recorded, so that the frequency and volume of the service personnel can be adjusted more efficiently when the customer conducts business again, further enhancing the customer's service experience.

10: Voice Service System 100: Voice Service Device 110, 210: Processor 120, 220: Storage Media 130, 230: Transceiver 141: First-Side Amplifier 142: Second-Side Amplifier 200: Server S210, S220, S230, S240, S250, S260, S270, S280, S290, S310, S320, S330, S340: Steps

Figure 1 is a schematic diagram of a voice service system with customer identification, according to an embodiment of the present invention. Figure 2 is a flowchart of a method applied to the voice service system with customer identification, according to an embodiment of the present invention. Figure 3 is a flowchart of a method applied to the voice service system with customer identification, according to an embodiment of the present invention.

10: Voice Service System

100: Voice Service Device

110, 210: Processor

120, 220: Storage Media

130, 230: Transceivers

141: First-side receiver/amplifier

142: Second-side receiver/amplifier

200: Server

Claims (10)

A voice service system with customer identification includes: a server for storing voice settings; and a voice service device including: a first-side receiver/amplifier and a second-side receiver/amplifier; a transceiver communicatively connected to the server; a storage medium; and a processor coupled to the transceiver, the storage medium, the first-side receiver/amplifier, and the second-side receiver/amplifier, wherein the processor is configured to perform: transmitting customer authentication information to the server to retrieve voice settings corresponding to the customer from the server; In response to identifying semantic keywords in a voice signal received from a service provider by the first-side receiver/amplifier, adjusting at least one of the frequency and volume values of the voice settings; adjusting the frequency and volume of the voice signal according to the voice settings; and transmitting the voice signal to the second-side receiver/amplifier. The voice service system as described in claim 1, wherein the processor is further configured to respond to the failure to identify semantic keywords from the voice signal without adjusting the voice settings. The voice service system as described in claim 1, wherein the processor is further configured to: in response to receiving a voice signal from the first-side receiver/amplifier, determine whether a semantic keyword is identified from the voice signal; and in response to identifying a semantic keyword from the voice signal, adjust at least one of the frequency value and volume value of the voice setting. The voice service system as described in claim 1, wherein the processor is further configured to respond to determining that the voice settings have been adjusted and to transmit the adjusted voice settings to the server. The voice service system of claim 1, wherein the server stores a machine learning model, and the processor is further configured to: recognize the semantic keywords from the voice signal, transmit the semantic keywords and the voice settings as inputs to the machine learning model to the server to obtain suggested values corresponding to the voice settings; and adjust at least one of the frequency value and the volume value of the voice settings according to the suggested values. The voice service system as described in claim 5, wherein the processor is further configured to associate all identified semantic keywords with the voice setting adjustment process as training data and transmit them to the server, wherein the server trains or updates the machine learning model based on the training data. The voice service system as claimed in claim 1, wherein the processor is further configured to, when transmitting the customer's authentication information to the server and the server notifies that there is no corresponding voice setting for the customer, set the frequency value and volume value obtained by analyzing the voice signal to the frequency value and volume value of the voice setting. The voice service system as claimed in claim 1, wherein the processor is further configured to denoise the voice signal before transmitting it to the second-side amplification device, and to denoise the voice signal received from the second-side amplification device before transmitting it to the first-side amplification device. The voice service system as described in claim 1, wherein the voice service system further includes an ATM, and the server is further configured to perform: When the customer operates the ATM, an identity verification procedure is performed on the customer through the ATM, and in response to successful identity verification, the corresponding voice settings for the customer are transmitted to the ATM. A voice service method with customer identification includes: transmitting customer identity verification information to a server to obtain voice settings corresponding to the customer from the server; identifying semantic keywords in response to a service personnel's voice signal received from a first-side receiver/amplifier, and adjusting at least one of the frequency and volume values of the voice settings; adjusting the frequency and volume of the voice signal according to the voice settings; and transmitting the voice signal to a second-side receiver/amplifier.