CN112667270A

CN112667270A - Voice processing resource updating method, computer equipment and storage device

Info

Publication number: CN112667270A
Application number: CN202011539620.4A
Authority: CN
Inventors: 舒文
Original assignee: iFlytek Co Ltd
Current assignee: iFlytek Co Ltd
Priority date: 2020-12-23
Filing date: 2020-12-23
Publication date: 2021-04-16
Anticipated expiration: 2040-12-23
Also published as: CN112667270B

Abstract

The application discloses a method for updating voice processing resources, computer equipment and a storage device. The method comprises the following steps: configuring a first storage unit into a service working state and configuring a second storage unit into an update working state, wherein the storage unit in the service working state is used for processing voice data by using voice processing resources, and the storage unit in the update working state is used for responding to a resource update request and executing the update of the voice processing resources; receiving a resource updating request, and updating the voice processing resource by using the second storage unit in the updating working state; and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state. By the scheme, the processing of the voice data can not be interrupted when the voice processing resources are updated.

Description

Voice processing resource updating method, computer equipment and storage device

Technical Field

The present application relates to the field of speech processing, and in particular, to a method for updating speech processing resources, a computer device, and a storage apparatus.

Background

As speech processing technology evolves, speech processing is increasingly applied in a variety of environments. However, for example, in the process of performing speech recognition, when it is necessary to update the speech recognition model, the speech recognition is usually performed first, and after the speech recognition model is updated, the speech recognition is performed again. This approach can interrupt the speech recognition process, resulting in problems such as stutter during the speech recognition process.

Disclosure of Invention

The technical problem mainly solved by the present application is to provide an updating method of voice processing resources, a computer device and a storage device, which can not interrupt the processing of voice data when updating the voice processing resources.

In order to solve the above problem, a first aspect of the present application provides an updating method for speech processing resources. The method comprises the following steps: configuring a first storage unit into a service working state and configuring a second storage unit into an update working state, wherein the storage unit in the service working state is used for processing voice data by using voice processing resources, and the storage unit in the update working state is used for responding to a resource update request and executing the update of the voice processing resources; receiving a resource updating request, and updating the voice processing resource by using the second storage unit in the updating working state; and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state.

In order to solve the above problem, a second aspect of the present application provides a computer device, which includes a memory and a processor coupled to each other, wherein the memory stores program data, and the processor is configured to execute the program data to implement any step of the above method for updating speech processing resources.

In order to solve the above problem, a third aspect of the present application provides a storage device storing program data capable of being executed by a processor, the program data being used to implement any one of the steps in the above method for updating speech processing resources.

In the above solution, the first storage unit is configured to be in a service working state, the second storage unit is configured to be in an update working state, the storage unit in the service working state is used for processing the voice data by using the voice processing resource, the storage unit in the update working state is used for performing the update of the voice processing resource in response to the resource update request, and the storage units in different working states can be used for respectively processing the update of the voice data and the voice processing resource; when a resource updating request is received, the second storage unit in the updating working state is used for updating the voice processing resource, after the voice processing resource is updated, the first storage unit is configured to be in the updating working state, and the second storage unit is configured to be in the service working state; the second storage unit is used for updating the voice processing resource, so that the voice data processing service can be continuously provided without interrupting the processing of the voice data by using the first storage unit when the voice processing resource is updated.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a flowchart illustrating a first embodiment of a method for updating speech processing resources according to the present application;

FIG. 2 is a flowchart illustrating a second embodiment of a method for updating speech processing resources according to the present application;

FIG. 3 is a flowchart illustrating a third embodiment of a method for updating speech processing resources according to the present application;

FIG. 4 is a flowchart illustrating a fourth embodiment of a method for updating speech processing resources according to the present application;

FIG. 5 is a schematic block diagram of an embodiment of a computer apparatus of the present application;

FIG. 6 is a schematic structural diagram of an embodiment of a memory device according to the present application.

Detailed Description

The following describes in detail the embodiments of the present application with reference to the drawings attached hereto.

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those skilled in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments without conflict.

The terms "system" and "network" are often used interchangeably herein. The term "and/or" herein is merely an association describing an associated object, meaning that three relationships may exist, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship. Further, the term "plurality" herein means two or more than two.

Referring to fig. 1, fig. 1 is a flowchart illustrating a first embodiment of a method for updating speech processing resources according to the present application. The method comprises the following steps:

s11: and configuring the first storage unit into a service working state, and configuring the second storage unit into an update working state.

The storage unit may be a memory block, a memory interval, or the like of the storage device. For example, one of the memory blocks of the storage device may be used as the first storage unit, and one memory block may be used as the second storage unit; the storage device may also be divided into a plurality of memory intervals, and one of the memory intervals is used as the first memory unit, and one of the memory intervals is used as the second memory unit, and the like, which is not limited in this application.

The operating states that may be configured for the memory cells may include a plurality of operating states, the operating states including at least one of: a service working state, an update working state, a transition working state, and the like. The storage unit in the service operating state is used for processing the voice data by using the voice processing resource, for example, the first storage unit in the service operating state may be used for performing voice recognition on the voice data such as the recording by using the voice recognition model so as to convert the voice data into information such as a text. The storage unit in the update operating state is used to perform updating of the speech processing resource, such as updating of a speech recognition model, in response to a resource update request. The transitional working state is an intermediate state between the service working state and the updating working state, and the storage unit in the working state is used for processing voice processing resources, voice data and the like in the updating working state and the state switching between the service working states.

In some embodiments, after receiving the service processing request, the storage unit in the service working state may be used to process the voice data; and after receiving the resource updating request, updating the voice processing resource by using the storage unit in the updating working state. In addition, the storage unit in the service working state may be used to receive a service processing request, and when the service processing request is received, the storage unit in the service working state is used to process voice data by using the voice processing resource. The resource update request may be received with the storage unit in an update-working state, and when the resource update request is received, the storage unit in the update-working state is made available to perform an update of the speech processing resource in response to the resource update request.

S12: and receiving a resource updating request, and updating the voice processing resource by using the second storage unit in the updating working state.

And if the resource updating request is received, updating the voice processing resource by using the second storage unit in the updating working state, and storing the voice processing resource updating data. For example, when an update request of the speech recognition model is received, the second storage unit in the update operation state may be used to load the update data of the speech recognition model for updating, and store the updated speech recognition model data.

Optionally, when a resource update request is received, the second storage unit is switched from the update operating state to the transition operating state, so that the second storage unit is no longer used to respond to the newly received resource update request, and thus the second storage unit in the transition operating state is used to update the voice processing resource.

S13: and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state.

After the updating of the voice processing resource by the second storage unit in the transitional working state is completed, the first storage unit is configured to be in the updating working state, so that a new resource updating request can be received by the first storage unit in the updating working state, and the updating of the voice processing resource is executed by the first storage unit in the updating working state in response to the new resource updating request.

After the second storage unit in the transition working state is used for updating the voice processing resource, the second storage unit is configured to be in the service working state, and it can be understood that the second storage unit can be switched from the transition working state to the service working state, so that the second storage unit in the service working state is used for receiving a new service processing request, and the second storage unit in the service working state is used for processing voice data by using the voice processing resource. The first storage unit after switching the working state may perform the working of the second storage unit before switching the working state, and the second storage unit after switching the working state may perform the working of the first storage unit before switching the working state.

Optionally, for step S13, configuring the first storage unit to update the operating state includes: and switching the first storage unit from the service working state to the updating working state so as to not process the received and unprocessed voice data by using the first storage unit. Specifically, if the service processing request is not received or the voice processing resource of the received service processing request has completed processing the voice data, the first storage unit may be switched from the working state to the updated working state, or the first storage may be switched from the working state to the transition working state and then switched to the updated working state, so that the first storage unit is no longer used to process the received and unprocessed voice data.

Optionally, if an unprocessed service processing request is received when the service processing request responded in the first storage unit is used to process the voice data, the received unprocessed voice data exists, and the received unprocessed voice data may be processed by using the voice processing resource in the second storage unit after the second storage unit is configured in the service working state. Specifically, the received and unprocessed voice data in the first memory list may be sent to the second memory unit, so that the received and unprocessed voice data is processed by using the voice processing resource of the second memory unit.

Optionally, regarding step S13, configuring the first storage unit to update the operating state may further include: and switching the first storage unit from the service working state to the transition working state so as to utilize the first storage unit in the transition working state to continue processing the received and unprocessed voice data, but not to utilize the first storage unit to respond to a new service processing request so as to not process the newly received voice data. After the received and unprocessed voice data is processed by the first storage unit, the first storage unit is switched from the transition working state to the updating working state, so that the updating of the voice processing resource is executed by the first storage unit in the updating working state in response to the resource updating request.

In this embodiment, the first storage unit is configured to be in a service operating state, the second storage unit is configured to be in an update operating state, the storage unit in the service operating state is used for processing voice data by using voice processing resources, the storage unit in the update operating state is used for executing update of the voice processing resources in response to a resource update request, and the storage units in different operating states can be used for respectively processing update of the voice data and the voice processing resources; when a resource updating request is received, the second storage unit in the updating working state is used for updating the voice processing resource, after the voice processing resource is updated, the first storage unit is configured to be in the updating working state, and the second storage unit is configured to be in the service working state; the second storage unit is used for updating the voice processing resource, so that the voice data processing service can be continuously provided without interrupting the processing of the voice data by using the first storage unit when the voice processing resource is updated.

Referring to fig. 2, fig. 2 is a flowchart illustrating a second embodiment of a method for updating speech processing resources according to the present application. The method comprises the following steps:

s21: and configuring the first storage unit into a service working state, and configuring the second storage unit into an update working state.

The first storage unit in the service working state stores a first voice processing resource and a second voice processing resource, the first voice processing resource can be a voice processing resource which is updated frequently or will be updated frequently, the second voice processing resource can be a voice processing resource which is updated infrequently or will not be updated frequently, and the combination of the first voice processing resource and the second voice processing resource is used for realizing the processing of voice data. At this time, the second storage unit in the update operation state may not store the voice processing resource. The first speech processing resource and the second speech processing resource are not limited by the present application.

When the resource updating request is received, the second storage unit in the updating working state can be used for updating the voice processing resource.

Optionally, the resource update request may be used to request to update the first voice processing resource, and the updated first voice processing resource is obtained by using the second storage unit in the update working state, so as to update the voice processing resource.

Optionally, the resource update request may be used to request to update the second speech processing resource, and the updated second speech processing resource is obtained by using the second storage unit in the update working state, so as to update the speech processing resource.

The present embodiment takes the resource update request as an example for requesting to update the first speech processing resource.

S22: and receiving a resource updating request, and acquiring the updated first voice processing resource by using the second storage unit in the updating working state.

After receiving the resource updating request, the second storage unit in the updating working state can be used for responding to the resource updating request, the second storage unit is switched to the transition working state, at the moment, the second storage unit is not used for responding to the new resource updating request, the second storage unit is used for loading the updating voice processing resource in the resource updating request so as to obtain the updated first voice processing resource, and the updated first voice processing resource is stored in the second storage unit.

S23: the second speech processing resource in the first storage unit is provided to the second storage unit.

Because the updated first voice processing resource is stored in the second storage unit, if the second storage unit is used for processing voice data by using the voice processing resource, and because the processing of the voice data is realized by using the combination of the first voice processing resource and the second voice processing resource, the second voice processing resource in the first storage unit needs to be provided to the second storage unit, so that the updated first voice processing resource and the second voice processing resource stored in the second storage unit are used for processing the voice data.

Optionally, the second voice processing resource in the first storage unit is sent to the second storage unit, so that the second storage unit can receive the second voice processing resource for storage. For example, the second speech processing resource in the first storage unit is moved to the second storage unit for storage.

Optionally, the second voice processing resource in the first storage unit is sent to the third storage unit for storage, so that the second storage unit can obtain the second voice processing resource by accessing the third storage unit. In some application scenarios, the third storage unit may be a memory block, a memory interval, and the like of the storage device, the first storage unit may mount the third storage unit, and after the second storage unit responds to the resource update request, the update condition of the voice processing resource may be sent to the first storage unit, where the update condition may include the updated second voice processing resource. In this case, the first storage unit may be switched from the service operating state to the transition operating state, and the second voice processing resource in the first storage unit may be transmitted to the third storage unit for storage. The first storage unit unloads the third storage unit, so that the second storage unit can mount the third storage unit, and the second voice processing resource stored in the third storage unit can be accessed through the second storage unit.

S24: and reserving or deleting the second voice processing resource in the first storage unit.

Optionally, if the first storage unit is switched to the transition working state, the first storage unit reserves the second speech processing resource, so as to process the received and unprocessed speech data by using the first speech processing resource and the second speech processing resource stored in the first storage unit. After the received and unprocessed voice data is processed, the first voice processing resource and the second voice processing resource in the first storage unit are deleted.

Optionally, if the first storage unit is switched to the transition working state, the second voice processing resource is deleted from the first storage unit, and the first storage unit and the second storage unit can mount the third storage unit at the same time to share the third storage unit, allowing the first storage unit to access the second voice processing resource in the third storage unit, so as to process the received and unprocessed voice data by using the first voice processing resource in the first storage unit and the second voice processing resource in the third storage unit. After the received and unprocessed voice data is processed, the first voice processing resource and the second voice processing resource in the first storage unit are deleted.

S25: and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state.

The first storage unit after switching the working state may perform the working of the second storage unit before switching the working state, and the second storage unit after switching the working state may perform the working of the first storage unit before switching the working state. The specific implementation of this embodiment can refer to the implementation process of the above embodiment, and is not described herein again.

In this embodiment, when receiving the resource update request, the second storage unit in the update working state is used to obtain the updated first voice processing resource, and then the second voice processing resource in the first storage unit is provided to the second storage unit, so that the second storage unit stores the second voice processing resource and the updated first voice processing resource, and after switching to the service working state, the updated voice processing resource can be used to continuously receive the new service processing request, thereby ensuring the sharing of the context data (voice processing resource) in the first storage unit and the second storage unit, and simultaneously providing the service processing service without interruption.

In some embodiments, for the above-mentioned embodiment step S24, after the received and unprocessed voice data is processed, if the second voice processing resource in the first storage unit is deleted and the first voice processing resource in the first storage unit that is not updated is deleted, after the voice processing resource is updated, the first storage unit is configured to be in the update operating state and the second storage unit is configured to be in the service operating state. And if a new service processing request is received, processing the voice data in the new service processing request by utilizing the updated first voice processing resource and the second voice processing resource in the second storage unit in the service working state. If a new resource update request is received, the first storage unit in the update working state is used to update the voice processing resource, which may specifically refer to the foregoing embodiments and will not be described herein again.

For step S24 of the above embodiment, after the received and unprocessed voice data is processed, if the first voice processing resource and the second voice processing resource in the first storage unit are reserved, after the voice processing resources are updated, the first storage unit is configured to be in the update operating state, and the second storage unit is configured to be in the service operating state. And if a new service processing request is received, processing the voice data in the new service processing request by utilizing the updated first voice processing resource and the second voice processing resource in the second storage unit in the service working state. And if a new resource updating request is received, updating the voice processing resource by using the first storage unit in the updating working state.

Optionally, if the resource update request is used to update the second speech processing resource, the first storage unit in the update working state may be used to respond to the resource update request, switch the first storage unit to the transition working state, and obtain the updated second speech processing resource by using the first storage unit in the update working state. At this time, the first voice processing resource which is stored in the second storage unit and is updated last time is sent to the first storage unit for storage. After the first storage unit is switched from the updated working state to the service working state, the first voice processing resource updated last time and the second voice processing resource updated this time are stored in the first storage unit to process a new service processing request.

Optionally, if the resource update request is used to update the first speech processing resource again, the first storage unit in the update working state may be used to respond to the resource update request, switch the first storage unit to the transition working state, and obtain the updated first speech processing resource by using the first storage unit in the update working state. In this case, the first storage unit stores the first speech processing resource and the second speech processing resource updated again, and the second speech processing resource in the second storage unit does not need to be transmitted to the first storage unit. After the first storage unit is switched from the updated working state to the service working state, the new service processing request can be processed by using the first voice processing resource and the second voice processing resource in the first storage unit.

In this embodiment, after the received and unprocessed voice data is processed, the second voice processing resource in the first storage unit is deleted, so that the memory space of the first storage unit can be saved, and the occupation of the storage space can be reduced. If the second voice processing resource in the first storage unit is reserved, the transmission operation of the voice processing resource between the first storage unit and the second storage unit can be reduced.

Referring to fig. 3, fig. 3 is a flowchart illustrating a third embodiment of a method for updating speech processing resources according to the present application. The method comprises the following steps:

s31: and configuring the first storage unit into a service working state, and configuring the second storage unit into an update working state.

S32: and receiving a resource updating request, and acquiring the updated first voice processing resource by using the second storage unit in the updating working state.

Optionally, when the resource update request is received and the second storage unit responds to the resource update request, if the resource update request is used to request to update the first voice processing resource, the updated first voice processing resource is obtained by using the second storage unit in the update working state; in addition, the second storage unit can be used for sending an operating state transition request to the first storage unit.

For the specific implementation of steps S31-S32 in this embodiment, reference may be made to the implementation processes of steps S11-S12 and S21-S22 in the above embodiment, which are not repeated herein.

S33: and presetting the second voice processing resource in the first storage unit to obtain a first verification identifier.

Optionally, the first storage unit is used to respond to the working state transition request, where the working state transition request carries an update condition of the voice processing resource, and the voice processing resource that is not updated in the first storage unit may be subjected to preset processing to obtain the first verification identifier.

Optionally, the second speech processing resource in the first storage unit is subjected to preset processing, where the preset processing may be hash operation or other tamper-resistant identification processing, and this application does not limit this. And performing hash operation on the second voice processing resource in the first storage unit to obtain a first hash value which can be used as the first verification identifier.

If the second voice processing resource comprises a plurality of voice processing resources or voice recognition models, the plurality of voice recognition models can be respectively subjected to Hash operation to obtain first sub-Hash values of the respective models, then the Hash operation is carried out on the sub-Hash values to obtain a first total Hash value, and the obtained first total Hash value and each first sub-Hash value can be used as a first verification identifier. For example, the second speech processing resource includes a plurality of speech recognition models, such as a main model, a session model, an object model, etc., and the models may be hashed:

h1 ═ Hash (H11 (master model) + H12 (session model) + H13 (object model)) (1)

Where H1 denotes a first total hash value, H11 denotes a first sub-hash value of the main model, H12 denotes a first sub-hash value of the session model, and H13 denotes a first sub-hash value of the object model.

S34: and sending the second voice processing resource in the first storage unit to a second storage unit.

For the specific implementation of this step, reference may be made to the implementation process of step S23 in the foregoing embodiment, which is not described herein again.

S35: and carrying out preset processing on the provided second voice processing resource to obtain a second verification identifier.

Performing preset processing on the received second voice processing resource by using a second storage unit, for example, performing hash operation on the second voice processing resource, and using an obtained second hash value as a second verification identifier; if the second voice processing resource comprises multiple voice processing resources, the multiple voice recognition models can be respectively subjected to hash operation to obtain second sub-hash values of the respective models, then the sub-hash values are subjected to hash operation to obtain second total hash values, and the obtained second total hash values and the second sub-hash values can be used as second verification identifiers. For example, the second speech processing resource includes a plurality of speech recognition models, such as a main model, a session model, an object model, etc., and the models may be hashed:

h2 ═ Hash (H21 (master model) + H22 (session model) + H23 (object model)) (2)

Where H2 denotes a second overall hash value, H21 denotes a second sub-hash value of the main model, H22 denotes a second sub-hash value of the session model, and H23 denotes a second sub-hash value of the object model.

S36: and judging whether the second verification identification is consistent with the first verification identification.

Whether the second hash value is equal to the first hash value or not can be judged, if so, the second verification identifier is judged to be consistent with the first verification identifier, and the second voice processing resource is not tampered and damaged; otherwise, the second verification identification is inconsistent with the first verification identification, and the second voice processing resource is tampered.

If the second verification identifier is consistent with the first verification identifier, performing step S37; if the second authentication identification is not consistent with the first authentication identification, step S38 is executed.

Optionally, if the first verification identifier includes the first total hash value and each first sub-hash value, and the second verification identifier includes the second total hash value and each second sub-hash value, it may be determined whether the second total hash value is equal to the first total hash value, and if so, it is determined that the second verification identifier is consistent with the first verification identifier, and step S37 is executed; if not, further determining whether each second sub-hash value is equal to each corresponding first sub-hash value to determine unequal sub-hash values, and performing step S38.

For example, the second voice processing resource includes a plurality of voice recognition models, such as a main model, a session model, an object model, etc., and after determining that the second total hash value H2 is not equal to the first total hash value H1, because the loading time of the main model is long, it may be further determined whether the second sub-hash value H21 of the main model is equal to the first sub-hash value H11 of the corresponding main model, if not, the main model is tampered with, and step S38 is executed; if the two types of the object model and the session model are equal, the main model is not tampered, and is not damaged, so that the main model cannot be reloaded, that is, the object model and/or the session model are tampered. It may be determined whether the second sub-hash values H22 of the session model, the second sub-hash values H23 of the object model and the corresponding first sub-hash values H12 and H13 are equal to determine unequal sub-hash values, that is, determine a tampered speech recognition model, and then step S38 is performed.

Optionally, after determining that the second total hash value H2 is not equal to the first total hash value H1, if it is determined that the second sub-hash value of the main model is equal to the first sub-hash value of the corresponding main model, indicating that the main model is not tampered, and a model is tampered in other models of the speech recognition model, then step S38 is performed.

S37: the second storage unit is allowed to obtain the provided second speech processing resources for the processing of the speech data.

Since the second voice processing resource received and stored by the second storage unit has not been tampered with, the second storage unit may be allowed to obtain the provided second voice processing resource for processing of the voice data.

S38: allowing the second storage unit to read at least a portion of the spare resources of the second voice processing resource from the fourth storage unit for processing of the voice data.

The fourth storage unit may be a memory block, a memory interval, or the like of the storage device, or may be a magnetic disk of the storage device, and the fourth storage unit, for example, the magnetic disk, stores the standby resources of all the voice processing resources. In addition, after the voice processing resource is updated, the updated voice processing resource can be stored in the fourth storage unit to update the standby resource.

And the second storage unit is used for storing the second voice processing resource received by the second storage unit, and the second voice processing resource received by the second storage unit is tampered with.

If the second speech processing resource comprises a plurality of speech processing resources or speech recognition models, the second storage unit is allowed to acquire the standby resource of the speech processing resource corresponding to the sub-hash value from the fourth storage unit according to the verification of the unequal sub-hash values.

For example, the second voice processing resource includes a plurality of voice recognition models, such as a main model, a session model, an object model, and the like, and when the total hash value is not equal, if it is determined that the sub-hash values of the main model are not equal and the main model is tampered with, the second storage unit is allowed to acquire a spare main model of the main model from the fourth storage unit; if the sub-hash values of the object model and/or the session model which are tampered are further judged to be not equal, the second storage unit is allowed to acquire the standby object model and/or session model from the fourth storage unit; if it is determined that the sub-hash values of the main model are equal, and the object model and/or the session model are loaded faster, the second storage unit may be allowed to acquire the standby object model and session model other than the main model from the fourth storage unit without determining whether the tampered sub-hash values of the object model and/or the session model are equal in the subsequent step.

S39: and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state.

After the voice processing resource is updated, if the first storage unit is switched to the transition working state, after the received and unprocessed voice data is processed by using the first voice processing resource and the second voice processing resource in the first storage unit, the second voice processing resource in the first storage unit can be reserved or deleted, and the first storage unit is configured to be in the update working state, so that the first storage unit is used for responding to a new resource update request.

And configuring the second storage unit into a service working state so as to process voice data by using the updated voice processing resources stored in the second storage unit and at least part of standby resources of the second voice processing resources read from the fourth storage unit.

In this embodiment, when updating the voice processing resource and sending the second voice processing resource in the first storage unit to the second storage unit, tamper-proof verification is performed to determine that the second voice processing resource stored in the second storage unit, which is not updated, is accurate and has not been tampered, so that the second storage unit is used to process the voice data in the new service processing request, and the accuracy of voice processing can be improved; if the second voice processing resource is verified to be tampered, the standby resource can be acquired to process the voice data, so that the voice data processing process is uninterrupted, and meanwhile, the safety of processing the voice data by using the first storage unit and the second storage unit can also be guaranteed; the tamper-proof verification can also prevent the voice data processing process from collapsing, the voice data processing speed is improved, and the first storage unit and the second storage unit complete the switching of the working state through hot switching, so that the switching speed can be improved; in addition, only the first storage unit and the second storage unit are needed in the data processing process, read-write separation is not needed, and the voice data processing efficiency can be improved.

Referring to fig. 4, fig. 4 is a flowchart illustrating a fourth embodiment of a method for updating speech processing resources according to the present application. The method comprises the following steps:

s40: and configuring the first storage unit into a service working state, and configuring the second storage unit into an update working state.

When the storage unit includes a first storage unit, a second storage unit, a third storage unit and a fourth storage unit, and the voice processing resource includes a first voice processing resource and a second voice processing resource, the first voice processing resource may be a voice processing resource that is updated frequently or will be updated frequently, and the second voice processing resource may be a voice processing resource that is updated infrequently or will not be updated frequently. The first storage unit may mount a third storage unit, and a combination of voice processing resources stored in the first storage unit and the third storage unit is used to implement processing of voice data. The voice processing resource condition stored in the storage unit may include at least one of the following.

Optionally, the first storage unit in the service working state stores a first voice processing resource and a second voice processing resource, and a combination of the first voice processing resource and the second voice processing resource is used to implement processing of voice data. At this time, the third storage unit and the second storage unit in the update operation state may not store the voice processing resource.

Optionally, the first storage unit in the service working state stores the second voice processing resource, the first storage unit mounts the third storage unit, and the third storage unit stores the first voice processing resource; in addition, the first storage unit can store a first voice processing resource, and the third storage unit stores a second voice processing resource; the combination of the first speech processing resource and the second speech processing resource is used to implement the processing of the speech data. At this time, the second storage unit in the update operation state may not store the voice processing resource.

Optionally, the first storage unit mounts a third storage unit, the third storage unit may store the first voice processing resource and the second voice processing resource, and a combination of the first voice processing resource and the second voice processing resource is used to implement processing of voice data. At this time, the second storage unit in the update operation state may not store the voice processing resource.

If the voice processing resources include a first voice processing resource, a second voice processing resource and a third voice processing resource, the update frequency of the voice processing resources can be counted, the voice processing resources with more total update frequency or more update frequency in the latest preset time period are used as the first voice processing resources, the voice processing resources with relatively moderate update frequency or infrequent update frequency are used as the second voice processing resources, and the voice processing resources with less update frequency or no update frequency are used as the third voice processing resources. The first storage unit may mount a third storage unit, where a combination of the voice processing resources stored in the first storage unit and the third storage unit is used to implement processing of voice data, and a situation of the voice processing resources stored in the storage unit is not limited in this application.

Optionally, the first storage unit stores a first voice processing resource and a second voice processing resource, the first storage unit mounts a third storage unit, the third storage unit stores a third voice processing resource, and a combination of the first voice processing resource, the second voice processing resource, and the third voice processing resource is used for implementing processing of voice data.

When receiving the service processing request, because the first storage unit mounts the third storage unit, the first storage unit can access the third voice processing resource stored in the third storage unit, and the first voice processing resource and the second voice processing resource in the first storage unit and the third voice processing resource in the third storage unit are utilized to process the voice data in the service processing request.

Optionally, the first storage unit stores a first voice processing resource, the first storage unit mounts a third storage unit, the third storage unit stores a second voice processing resource and a third voice processing resource, and a combination of the first voice processing resource, the second voice processing resource and the third voice processing resource is used for implementing processing of voice data.

Optionally, the first storage unit may not store the voice processing resource, the first storage unit mounts a third storage unit, the third storage unit stores the first voice processing resource, the second voice processing resource, and the third voice processing resource, and a combination of the first voice processing resource, the second voice processing resource, and the third voice processing resource is used to implement processing of the voice data.

In this embodiment, a first storage unit stores a first voice processing resource and a second voice processing resource, and a third storage unit stores a third voice processing resource.

S41: and receiving a resource updating request, and updating the voice processing resource by using the second storage unit in the updating working state.

When receiving the resource updating request, the second storage unit responds to the resource updating request, the second storage unit is switched from the updating working state to the transition working state, and the second storage unit in the updating working state can be used for updating the voice processing resource. In addition, the second storage unit can be used for sending an operating state transition request to the first storage unit.

Optionally, the resource update request is used to request to update the first speech processing resource, and the updated first speech processing resource may be obtained by using the second storage unit in the update working state, so as to update the speech processing resource.

Optionally, the resource update request may also be used to request to update the second speech processing resource, and the updated second speech processing resource may be obtained by using the second storage unit in the update operating state.

Optionally, the resource update request may also be used to request to update the third speech processing resource, and the updated third speech processing resource may be obtained by using the second storage unit in the update operating state.

S42: and presetting the second voice processing resource in the first storage unit and the third voice processing resource in the third storage unit to obtain a third verification identifier.

Optionally, the voice processing resource not updated in the first storage unit and the third storage unit is subjected to preset processing, so as to obtain a third verification identifier. Because the resource update request is used for requesting to update the first voice processing resource, the second voice processing resource and the third voice processing resource are not updated, and the second voice processing resource in the first storage unit and the third voice processing resource in the third storage unit can be subjected to preset processing to obtain the third verification identifier.

Optionally, after receiving the working state switching request, the first storage unit switches the first storage unit from the service working state to the transition working state, and may acquire the object lock of the third storage unit by using the first storage unit, package and mount the second voice processing resource in the first storage unit into the third storage unit for storage, and simultaneously seal and store the third voice processing resource stored in the third storage unit. And performing hash operation on the second voice processing resource and the third voice processing resource in the third storage unit to obtain a hash value serving as a third verification identifier, and storing the third verification identifier in the third storage unit. The acquired object lock of the third storage unit may then be released. And acquiring the object lock of the third storage unit, and operating the third storage unit.

S43: the second speech processing resource in the first storage unit is provided to the second storage unit.

Optionally, the second voice processing resource in the first storage unit is sent to the third storage unit for storage, so that the second storage unit can obtain the second voice processing resource by accessing the third storage unit. Specifically, due to the update of the first voice processing resource, the second voice processing resource and the third voice processing resource for voice processing are lacked in the second storage unit, and the second voice processing resource in the first storage unit can be sent to the third storage unit for storage. The third storage unit is unloaded from the first storage unit, so that the second storage unit can mount the third storage unit, so that the second storage unit can obtain the second voice processing resource by accessing the third storage unit, and the third voice processing resource in the third storage unit is used for voice data processing.

Optionally, when the second voice processing resource in the first storage unit is sent to the third storage unit for storage, the second voice processing resource is simultaneously reserved in the first storage unit, and the third voice processing resource in the third storage unit is sent to the first storage unit for storage, so that the first voice processing resource, the second voice processing resource, and the third voice processing resource are stored in the first storage unit, and the voice data can be processed by using the voice processing resource of the first storage unit.

S44: and presetting the provided second voice processing resource and a third voice processing resource in a third storage unit to obtain a fourth verification identifier.

The second storage unit can be used for acquiring the object lock of the third storage unit, and preset processing, such as hash operation, is performed on the second voice processing resource in the third storage unit and the third voice processing resource in the third storage unit, and the obtained hash value can be used as the fourth verification identifier.

S45: and judging whether the fourth verification identification is consistent with the third verification identification.

If the fourth authentication identifier is determined to be consistent with the third authentication identifier, executing step S46; if it is determined that the fourth authentication flag does not match the third authentication flag, the step S47 is executed.

S46: the second storage unit is allowed to access the provided second speech processing resource and a third speech processing resource in a third storage unit for processing of speech data.

And the second storage unit can be allowed to access the third storage unit to read the second voice processing resource and the third voice processing resource for processing the voice data due to the fact that the second voice processing resource and the third voice processing resource are not tampered.

S47: allowing the second storage unit to read at least part of the spare resources of the second speech processing resource and/or the spare resources of the third speech processing resource from the fourth storage unit for processing of the speech data.

If it is verified that the second voice processing resource is tampered, or if it is verified that a part of the voice processing resource of the second voice processing resource is tampered, the second storage unit may be allowed to read all or a part of the spare resources of the second voice processing resource from the fourth storage unit. If it is verified that the third voice processing resource is tampered, or if it is verified that a part of the third voice processing resource is tampered, the second storage unit may be allowed to read all or a part of the spare resources of the third voice processing resource from the fourth storage unit. If the second voice processing resource and the third voice processing resource are verified to be tampered, the second storage unit can be allowed to read the standby resources of the second voice processing resource and the third voice processing resource from the fourth storage unit. That is, verifying the tampered voice processing resource may allow the second storage unit to read the spare resource of the tampered voice processing resource from the fourth storage unit for processing of the voice data.

Alternatively, after the step S47, if the first storage unit is switched to the transition operating state, the third voice processing resource in the third storage unit may be stored in the first storage unit to process the received and unprocessed voice data by using the first voice processing resource, the second voice processing resource and the third voice processing resource in the first storage unit, or the first storage unit may be allowed to access the third voice processing resource in the third storage unit to process the received and unprocessed voice data by using the first voice processing resource, the second voice processing resource and the third voice processing resource in the third storage unit. After the received and unprocessed voice data is processed, the first voice processing resource, the second voice processing resource and the third voice processing resource in the first storage unit are deleted.

Optionally, after the step S47, after the received and unprocessed voice data is processed, the second voice processing resource and the third voice processing resource in the first storage unit are reserved, or the second voice processing resource in the first storage unit is reserved, so that when the voice processing resources are updated again later, the amount of movement of the voice processing resources can be reduced.

S48: and after the voice processing resource is updated, configuring the first storage unit into an update working state, and configuring the second storage unit into a service working state.

After the voice processing resource is updated, the second storage unit can be switched from the transition working state to the service working state; after the received and unprocessed voice data is processed by the first storage unit, the first storage unit can be switched from the transition working state to the update working state. The first storage unit after switching the working state may perform the working of the second storage unit before switching the working state, and the second storage unit after switching the working state may perform the working of the first storage unit before switching the working state.

S49: and receiving a new service processing request, and processing the voice data in the new service processing request by using the first voice processing resource and the second voice processing resource in the second storage unit and the third voice processing resource in the third storage unit.

If a new service processing request is received, the first voice processing resource and the second voice processing resource in the second storage unit and the third voice processing resource in the third storage unit can be used to process the voice data in the new service processing request.

Optionally, because the updated first voice processing resource is stored in the second storage unit, the third storage unit stores the second voice processing resource and the third voice processing resource, and the second storage unit mounts the third storage unit, the second voice processing resource and the third voice processing resource in the third storage unit are accessed by using the first voice processing resource in the second storage unit and the second storage unit, so as to process the voice data in the new service processing request.

After step S48, when a new resource update request is received, the speech processing resource is updated using the first storage unit in the update operation state. Specifically, the first storage unit can be used to respond to the resource update request, the first storage unit is switched from the update working state to the transition working state, and the first storage unit is used to update the voice processing resource.

Different from the above embodiments, in this embodiment, the third storage unit is used to store the third speech processing resource, and the first storage unit and the second storage unit can share the third speech processing resource in the third storage unit, so that the movement of the third speech processing resource can be reduced; meanwhile, the storage space of the first storage unit and the second storage unit for storing the voice processing resources is reduced, and the storage space of the first storage unit and the second storage unit is saved; in addition, the three storage units are used for processing and updating voice data, and read-write separation can be realized.

With reference to fig. 5, fig. 5 is a schematic structural diagram of an embodiment of a computer device according to the present application. The computer device 100 comprises a memory 101 and a processor 102, wherein the memory 101 and the processor 102 are coupled to each other, the memory 101 stores program data, and the processor 102 is configured to execute the program data to implement the steps of any of the above-described methods for updating speech processing resources.

In the present embodiment, the processor 102 may also be referred to as a CPU (Central Processing Unit). The processor 102 may be an integrated circuit chip having signal processing capabilities. The processor 102 may also be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. A general purpose processor may be a microprocessor or the processor 102 may be any conventional processor or the like.

The specific implementation of this embodiment can refer to the implementation process of the above embodiment, and is not described herein again.

For the method of the above embodiment, it can be implemented in the form of a computer program, so that the present application provides a storage device, please refer to fig. 6, where fig. 6 is a schematic structural diagram of an embodiment of the storage device of the present application. The storage device 200 stores therein program data 201 executable by a processor, the program data being executable by the processor to implement the steps of any of the above-described embodiments of the method for updating speech processing resources.

The storage device 200 of this embodiment may be a medium that can store program data, such as a usb disk, a portable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, or may be a server that stores the program data, and the server may transmit the stored program data to another device for operation, or may self-operate the stored program data.

In the several embodiments provided in the present application, it should be understood that the disclosed method and apparatus may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a module or a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some interfaces, and may be in an electrical, mechanical or other form.

Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a separate product, may be stored in a storage device, which is a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or contributed to by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, a network device, or the like) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present application.

It will be apparent to those skilled in the art that the modules or steps of the present application described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present application is not limited to any specific combination of hardware and software.

The above description is only for the purpose of illustrating embodiments of the present application and is not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application or are directly or indirectly applied to other related technical fields, are also included in the scope of the present application.

Claims

1. A method for updating speech processing resources, the method comprising:

configuring a first storage unit into a service working state and configuring a second storage unit into an update working state, wherein the storage unit in the service working state is used for processing voice data by using voice processing resources, and the storage unit in the update working state is used for responding to a resource update request and executing the update of the voice processing resources;

receiving a resource updating request, and updating the voice processing resource by using the second storage unit in the updating working state;

and after the voice processing resource is updated, configuring the first storage unit into the updating working state, and configuring the second storage unit into the service working state.

2. The method according to claim 1, wherein said updating the speech processing resource using the second storage unit in the update operating state comprises:

switching the second storage unit from the updating working state to a transition working state so as to not respond to the newly received resource updating request by the second storage unit;

updating the voice processing resource by utilizing the second storage unit in the transition working state;

the configuring the second storage unit to the service operating state includes:

and switching the second storage unit from the transition working state to the service working state.

3. The method of claim 1, wherein said configuring said first storage unit to said updated operational state comprises:

switching the first storage unit from the service working state to the updating working state so as to not process the received and unprocessed voice data by utilizing the first storage unit; alternatively, the first and second electrodes may be,

switching the first storage unit from the service working state to a transition working state so as to utilize the first storage unit to continuously process the received and unprocessed voice data but not to process the newly received voice data any more; and after the received and unprocessed voice data is processed by the first storage unit, switching the first storage unit from the transition working state to the updating working state.

4. The method of claim 3, wherein after the switching the first storage unit from the business operating state to the update operating state, the method further comprises:

and after the second storage unit is configured to be in the service working state, processing the received and unprocessed voice data by utilizing the voice processing resource in the second storage unit.

5. The method of claim 1, wherein the first storage unit stores a first voice processing resource and a second voice processing resource, wherein the resource update request requests an update of the first voice processing resource, and wherein the combination of the first voice processing resource and the second voice processing resource is used to implement the processing of the voice data;

the updating the voice processing resource by using the second storage unit in the updating working state comprises:

obtaining the updated first voice processing resource by utilizing a second storage unit in the updating working state;

after the voice processing resource is updated and before the second storage unit is configured to the service working state, the method further includes:

providing the second speech processing resource in the first storage unit to the second storage unit;

reserving or deleting the second voice processing resource in the first storage unit.

6. The method of claim 5, wherein providing the second speech processing resource in the first storage unit to the second storage unit comprises:

sending the second voice processing resource in the first storage unit to the second storage unit; alternatively, the first and second electrodes may be,

and sending the second voice processing resource in the first storage unit to a third storage unit for storage, so that the second storage unit can obtain the second voice processing resource by accessing the third storage unit.

7. The method of claim 6, wherein prior to said providing the second speech processing resource in the first storage unit to the second storage unit, the method further comprises:

presetting the second voice processing resource in the first storage unit to obtain a first verification identifier;

after the providing the second speech processing resource in the first storage unit to the second storage unit, the method further comprises:

performing the preset processing on the provided second voice processing resource to obtain a second verification identifier;

if the second verification identification is consistent with the first verification identification, allowing the second storage unit to obtain the provided second voice processing resource for processing the voice data;

and if the second verification identifier is inconsistent with the first verification identifier, allowing the second storage unit to read at least part of standby resources of the second voice processing resources from a fourth storage unit for processing the voice data.

8. The method of claim 6, wherein the third storage unit further comprises a third speech processing resource, and wherein a combination of the first speech processing resource, the second speech processing resource, and the third speech processing resource is used to implement the processing of the speech data; after configuring the second storage unit to the business operating state, the method further comprises:

and receiving a new service processing request, and processing the voice data in the new service processing request by using the first voice processing resource and the second voice processing resource in the second storage unit and the third voice processing resource in the third storage unit.

9. The method of claim 8, wherein prior to said providing the non-updated resources of the first storage unit to the second storage unit, the method further comprises:

presetting the non-updated resources to obtain a first verification identifier, wherein before the second voice processing resources in the first storage unit are provided to the second storage unit, the method further comprises:

presetting the second voice processing resource in the first storage unit and a third voice processing resource in the third storage unit to obtain a third verification identifier;

performing preset processing on the provided second voice processing resource and a third voice processing resource in a third storage unit to obtain a fourth verification identifier;

if the fourth verification identifier is consistent with the third verification identifier, allowing the second storage unit to access the provided second voice processing resource and a third voice processing resource in the third storage unit for processing the voice data;

and if the fourth verification identifier is inconsistent with the third verification identifier, allowing the second storage unit to read at least part of the standby resource of the second voice processing resource and/or the standby resource of the third voice processing resource from a fourth storage unit for processing the voice data.

10. The method of claim 8, wherein the deleting the second speech processing resource from the first storage unit comprises:

if the first storage unit is switched to a transition working state, storing a third voice processing resource in the third storage unit into the first storage unit so as to utilize the first voice processing resource, the second voice processing resource and the third voice processing resource in the first storage unit to process received and unprocessed voice data, or allowing the first storage unit to access the third voice processing resource in the third storage unit so as to utilize the first voice processing resource, the second voice processing resource and the third voice processing resource in the first storage unit to process the received and unprocessed voice data;

deleting the first voice processing resource, the second voice processing resource and the third voice processing resource in the first storage unit after the processing completes the received and unprocessed voice data.

11. A computer device comprising a memory and a processor coupled to each other, the memory having stored therein program data for execution by the processor to perform the steps of the method of any one of claims 1 to 10.

12. A storage device, characterized by program data stored therein which can be executed by a processor for carrying out the steps of the method according to any one of claims 1 to 10.