Semantic communication method and apparatus, and terminal, server, system and storage medium

Abstract

The present application relates to a semantic communication method and apparatus, and a terminal, a server, a system and a storage medium. The method comprises: a first terminal performing semantic processing on semantic communication information in a semantic communication process; and if the semantic processing fails, the first terminal sending the semantic communication information to a server, such that the server processes the semantic communication information, obtains semantic information corresponding to the semantic communication information, and returns the semantic information. By using the method, the communication reliability of semantic communication can be improved.

Description

Semantic communication method, apparatus, terminal, server, system and storage medium

[0001] Related applications

[0002] The present application claims priority to the Chinese patent application No. 202410832278.9, filed on June 26, 2024, entitled "Semantic communication method, apparatus, terminal, server, system and storage medium", the contents of which are hereby incorporated by reference in its entirety. TECHNICAL FIELD

[0003] The present application relates to the technical field of wireless communication, and in particular to a semantic communication method, apparatus, terminal, server, system and storage medium. BACKGROUND

[0004] With the rapid development of cloud computing, big data, Internet of Things, artificial intelligence and other technologies, the growth trend of information (data volume) is becoming more and more significant. In the face of massive information transmission demand, a semantic communication mode appears to solve the problem of limited transmission bandwidth.

[0005] In semantic communication, both the sending end and the receiving end have a background knowledge base, so as to realize semantic transmission between the two parties. However, the current semantic communication mode has the problem of poor communication reliability.

[0006] SUMMARY

[0007] The embodiments of the present application provide a semantic communication method, apparatus, terminal, server, system and storage medium, which can improve the communication reliability of semantic communication.

[0008] In a first aspect, the present application provides a semantic communication method, the method comprising: performing semantic processing on semantic communication information in a semantic communication process; if the semantic processing fails, sending the semantic communication information to a server to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and returning the semantic information.

[0009] In some embodiments, the semantic processing includes semantic extraction and / or semantic restoration.

[0010] In some embodiments, the semantic processing on the semantic communication information in the semantic communication process includes performing semantic processing on the semantic communication information by using a local semantic knowledge base model.

[0011] In some embodiments, the local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0012] In some embodiments, the method further includes receiving the semantic communication information sent by the second terminal.

[0013] In some embodiments, the method further includes receiving the semantic information sent by the server; wherein the semantic information is obtained by the server performing semantic processing on the semantic communication information through a global model, or the semantic information is obtained by the server performing semantic processing on the approximate feature of the semantic communication information through the global model.

[0014] In some embodiments, the method further includes obtaining the semantic information if the semantic processing is successful.

[0015] In some embodiments, the method further includes determining whether the current semantic communication process is completed according to the semantic information, and sending semantic communication completion information to the second terminal if it is determined that the current semantic communication process is completed.

[0016] In some embodiments, after the semantic communication completion information is sent to the second terminal, the method further includes performing an operation corresponding to the semantic information.

[0017] In some embodiments, the determining whether the current semantic communication process is completed according to the semantic information includes determining whether the semantic information meets a communication intent or a communication target corresponding to the semantic communication information, and determining that the current semantic communication process is completed if it is determined that the semantic information meets the communication intent or the communication target.

[0018] In some embodiments, the method further includes sending feedback information to the second terminal if it is determined that the current semantic communication process is not completed, wherein the feedback information is used to instruct the second terminal to re-perform the semantic communication process.

[0019] In some embodiments, the method further includes receiving the semantic communication information re-sent by the second terminal, and performing semantic processing on the semantic communication information re-sent by the second terminal to obtain the semantic information.

[0020] In some embodiments, the method further includes receiving updated model parameters sent by the server, wherein the updated model parameters are obtained by the server performing model parameter updating on the global model according to the approximate feature after obtaining the semantic information by performing semantic processing on the approximate feature, and updating the local semantic knowledge base model based on the updated model parameters.

[0021] In some embodiments, the method further comprises: receiving semantic processing failure feedback sent by the server; and sending the semantic processing failure feedback to a second terminal.

[0022] In a second aspect, the present application provides a semantic communication method. The method comprises: receiving semantic communication information sent by a first terminal, and performing semantic processing on the semantic communication information; and sending semantic information obtained by the semantic processing to the first terminal.

[0023] In some embodiments, the semantic processing on the semantic communication information comprises: performing semantic processing on the semantic communication information according to a global model; and if semantic information is not obtained, performing information decomposition on the semantic communication information to obtain approximate features, and performing semantic processing on the approximate features by using the global model.

[0024] In some embodiments, the method further comprises: receiving semantic communication completion information sent by a second terminal, wherein the semantic communication completion information is determined by the first terminal according to semantic information obtained by performing semantic processing on the approximate features by using a global model of a server; updating model parameters of the global model according to the approximate features to obtain updated model parameters; and sending the updated model parameters to the first terminal.

[0025] In some embodiments, the method further comprises: if semantic information is not obtained in the process of performing semantic processing on the approximate features by using the global model, sending semantic processing failure feedback to the first terminal.

[0026] In some embodiments, the method further comprises: if a plurality of failure notifications sent by the second terminal are received, deleting the approximate features obtained by the decomposition.

[0027] In a third aspect, the present application provides a semantic communication device. The device comprises: a semantic processing module configured to perform semantic processing on semantic communication information in a semantic communication process; and a sending module configured to, if the semantic processing fails, send the semantic communication information to a server, so that the server processes the semantic communication information and obtains semantic information corresponding to the semantic communication information, and returns the semantic information.

[0028] In a fourth aspect, the present application provides a semantic communication device. The device comprises: a receiving module configured to receive semantic communication information sent by a first terminal, and perform semantic processing on the semantic communication information; and a sending module configured to send semantic information obtained by the semantic processing to the first terminal.

[0029] The application provides a terminal in the fifth aspect. The terminal comprises a transmitter and a processor; the processor is used for performing semantic processing on semantic communication information in a semantic communication process; and the transmitter is used for sending the semantic communication information to a server if the semantic processing fails, so as to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and return the semantic information.

[0030] The application provides a server in the sixth aspect. The server comprises a transmitter, a processor and a receiver; the receiver is used for receiving semantic communication information sent by a first terminal; the processor is used for performing semantic processing on the semantic communication information; and the transmitter is used for sending semantic information obtained by the semantic processing to the first terminal.

[0031] The application provides a semantic communication system in the seventh aspect. The semantic communication system comprises a first terminal, a second terminal and a server; the second terminal is used for sending semantic communication information to the first terminal; the first terminal is used for performing the steps of the method in the first aspect; and the server is used for performing the steps of the method in the second aspect.

[0032] In some embodiments, the second terminal is further used for performing knowledge base searching according to the to-be-transmitted communication information, so as to obtain the semantic communication information; and the second terminal is further used for sending the to-be-transmitted communication information to the server if the knowledge base searching fails, so as to perform knowledge base searching according to the to-be-transmitted communication information by the server, so as to obtain the semantic communication information.

[0033] The application provides a computer readable storage medium in the eighth aspect. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to realize the steps of the method in the first aspect or the second aspect.

[0034] The application provides a chip in the ninth aspect. The chip comprises a programmable logic circuit and / or program instructions, and realizes the steps of the method in the first aspect or the second aspect when the chip is running.

[0035] The application provides a computer program product in the tenth aspect. The computer program product comprises a computer program, and the computer program is executed by a processor to realize the steps of the method in the first aspect or the second aspect.

[0036] The semantic communication method, apparatus, terminal, server, system and storage medium described above, the first terminal performs semantic processing on the semantic communication information in the semantic communication process, if the semantic processing fails, the first terminal sends the semantic communication information to the server, so that the server processes the semantic communication information and obtains the semantic information corresponding to the semantic communication information, and returns the semantic information, compared with the way that the sending end and the receiving end only rely on the local background knowledge base for semantic analysis in the prior art, in the case that the first terminal is insufficient in capability, i.e. the first terminal fails to successfully perform semantic processing on the semantic communication information, the first terminal can borrow the capability of the server, perform semantic processing on the semantic communication information through the server, and then the server returns the obtained semantic information to the first terminal, thereby improving the success rate of semantic processing, ensuring the smooth progress of the semantic communication process, and improving the communication reliability of semantic communication. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the following will briefly introduce the drawings needed to be used in the embodiments or the related art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0038] FIG. 1 is an implementation environment diagram of a semantic communication method in an embodiment of the present application.

[0039] FIG. 2 is a flowchart of a semantic communication method in an embodiment of the present application.

[0040] FIG. 3 is a flowchart of a semantic communication method in another embodiment of the present application.

[0041] FIG. 4 is a flowchart of a semantic communication method in another embodiment of the present application.

[0042] FIG. 5 is a flowchart of the first terminal determining whether the semantic communication process is completed in another embodiment of the present application.

[0043] FIG. 6 is a flowchart of a semantic communication method in another embodiment of the present application.

[0044] FIG. 7 is a flowchart of a semantic communication method in another embodiment of the present application.

[0045] FIG. 8 is a flowchart of a semantic communication method in another embodiment of the present application.

[0046] FIG. 9 is a flowchart of a semantic communication method in another embodiment of the present application.

[0047] FIG. 10 is a flowchart of a semantic communication method in another embodiment of the present application.

[0048] Fig. 11 is a structural block diagram of a semantic communication device according to an embodiment of the present application.

[0049] Fig. 12 is a structural block diagram of a semantic communication device according to another embodiment of the present application.

[0050] Fig. 13 is a structural schematic diagram of a terminal according to an embodiment of the present application.

[0051] Fig. 14 is a structural schematic diagram of a server according to an embodiment of the present application.

[0052] Fig. 15 is a schematic structural diagram of a chip according to an embodiment of the present application. DETAILED DESCRIPTION

[0053] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0054] With the rapid development of cloud computing, big data, Internet of Things, artificial intelligence and other technologies, the growth trend of information (data volume) is becoming more and more significant. In the face of massive information transmission demand, a semantic communication mode has emerged to solve the problem of limited transmission bandwidth.

[0055] Semantic communication can integrate user's demand and information meaning into the communication process. The fundamental purpose of semantic communication is information exchange. It does not pursue the fidelity of original data or signal, but pursues the accurate transmission of semantics between the two parties. Therefore, in semantic communication, the sending end and the receiving end have background knowledge bases. The two parties communicate according to the background knowledge base to realize semantic transmission between the two parties. According to the difference of the background knowledge base, point-to-point semantic communication can be divided into three categories: semantic communication based on knowledge graph, semantic communication based on semantic base, and semantic communication based on basic model.

[0056] However, the current semantic communication mode has the problem of poor communication reliability. For example, the sending end and the receiving end cannot successfully perform semantic analysis according to their respective background knowledge bases (such as the receiving end cannot successfully perform semantic restoration according to the local background knowledge base), which will lead to failure of semantic communication, and further lead to poor communication reliability of semantic communication.

[0057] In view of this, the embodiments of the present application provide a semantic communication method, device, terminal, server, system and storage medium, which can improve the communication reliability of semantic communication.

[0058] It should be noted that the beneficial effects brought by the embodiments of the present application or the technical problems solved are not limited to this, but also other implicit or related problems, which can be seen from the description of the following embodiments.

[0059] Fig. 1 is a schematic diagram of an application environment of a semantic communication method provided by an embodiment of the present application. In the figure, a first terminal 102 communicates with a second terminal 104 through a network, the first terminal 102 communicates with a server 106 through the network, and the second terminal 104 communicates with a server 108 through the network.

[0060] The first terminal 102 and the second terminal 104 can be, but are not limited to, various personal computers, notebook computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things device can be a smart speaker, a smart television, a smart air conditioner, a smart vehicle-mounted device, a projection device, etc. The portable wearable device can be a smart watch, a smart bracelet, a head-mounted device, etc. The head-mounted device can be a virtual reality (VR) device, an augmented reality (AR) device, smart glasses, etc.

[0061] The server 106 and the server 108 can be the same server or different servers. Taking the server 106 as an example, the server 106 can be a standalone physical server, or a server cluster or distributed system composed of multiple physical servers, or an edge server, a cloud server providing cloud computing services, etc. The server 108 is similar to the server 106, and will not be described here.

[0062] In an embodiment of the present application, as shown in Fig. 2, a semantic communication method is provided. Taking the first terminal in Fig. 1 as an example, the method includes the following steps S201 to S202.

[0063] Step S201: The first terminal performs semantic processing on semantic communication information in a semantic communication process.

[0064] The first terminal and the second terminal perform semantic communication. Specifically, the first terminal and the second terminal can be heterogeneous terminals, for example, the computing power, hardware capability, power, etc. of the first terminal and the second terminal are different; alternatively, the related parameters of the first terminal and the second terminal can also be the same, which is not limited here.

[0065] In the embodiments of the present application, the number of first terminals can be one or more, and the number of second terminals can also be one or more. For the convenience of description, the number of first terminals and the number of second terminals are both taken as one in the following embodiments.

[0066] In the semantic communication process, the first terminal can be a sending terminal or a receiving terminal. The sending terminal refers to a transmitting party in the semantic communication process, which is configured to modulate the sending information. The receiving terminal refers to a receiving party in the semantic communication process, which is configured to demodulate the receiving information. Taking a one-way information transmission process in the semantic communication process as an example, the sending terminal extracts semantics from the communication information to be transmitted, so as to extract semantic features from the communication information to be transmitted. The sending terminal sends the semantic features to the receiving terminal. The receiving terminal restores the semantic features after receiving the semantic features, and restores the communication information.

[0067] In the embodiment of the application, the semantic processing includes semantic extraction and / or semantic restoration. In the case that the first terminal is a sending terminal, the semantic communication information can be the communication information to be transmitted obtained by the first terminal. The semantic processing of the first terminal on the semantic communication information can mean that the first terminal extracts semantics from the semantic communication information to obtain semantic information corresponding to the semantic communication information. The semantic information can be semantic features extracted from the semantic communication information.

[0068] In the case that the first terminal is a receiving terminal, the first terminal receives the semantic communication information sent by the second terminal. The semantic communication information can be semantic features sent by the second terminal. The semantic processing of the first terminal on the semantic communication information can mean that the first terminal restores the semantic communication information to obtain semantic information corresponding to the semantic communication information. The semantic information can be complete communication information restored from the semantic features.

[0069] In the embodiment of the application, the first terminal can use a knowledge graph to process the semantic communication information and attempt to obtain the semantic information corresponding to the semantic communication information. Optionally, the first terminal can also use a semantic base to process the semantic communication information and attempt to obtain the semantic information corresponding to the semantic communication information. The way in which the first terminal processes the semantic communication information will be exemplarily described in the following embodiments.

[0070] Step S202: If the semantic processing fails, the first terminal sends the semantic communication information to the server.

[0071] The semantic communication information is used for processing by the server to obtain the semantic information corresponding to the semantic communication information, and returning the semantic information.

[0072] The first terminal processes the semantic communication information in the manner of step S201. If the semantic processing is successful, the first terminal obtains the semantic information corresponding to the semantic communication information.

[0073] If the semantic processing fails, that is, the first terminal does not successfully obtain the semantic information corresponding to the semantic communication information, the first terminal sends the semantic communication information to the server.

[0074] After the server receives the semantic communication information sent by the first terminal, the server performs semantic processing on the semantic communication information, and then sends the semantic information obtained by the semantic processing to the first terminal. The process of the server performing semantic processing on the semantic communication information will be described in the following embodiments.

[0075] In this way, in the case where the processing capability of the first terminal is insufficient, i.e., the first terminal fails to successfully perform semantic processing on the semantic communication information, the first terminal can borrow the capability of the server to perform semantic processing on the semantic communication information through the server, and the server returns the obtained semantic information to the first terminal. In this way, the success rate of semantic processing can be improved, and the communication reliability of semantic communication can also be improved.

[0076] It should be noted that in the case where the first terminal is a receiving terminal and the second terminal is a sending terminal, in addition to the first terminal borrowing the processing capability of the server to perform semantic processing on the semantic communication information, the second terminal can also borrow the processing capability of the server. When the second terminal obtains the semantic communication information, the second terminal first performs knowledge base searching according to the to-be-transmitted communication information to attempt to obtain the semantic communication information. If the knowledge base searching fails, the second terminal sends the to-be-transmitted communication information to the server, and the server performs knowledge base searching according to the to-be-transmitted communication information to obtain the semantic communication information. In this way, the smooth progress of the semantic communication process can be ensured.

[0077] In an embodiment, based on the embodiment shown in FIG. 2 and in combination with the embodiment shown in FIG. 3, this embodiment will exemplarily introduce a manner in which the first terminal performs semantic processing on the semantic communication information. As shown in FIG. 3, in this embodiment, step S201 includes step S2011 shown in FIG. 3.

[0078] Step S2011: The first terminal performs semantic processing on the semantic communication information by using the local semantic knowledge base model stored in the first terminal.

[0079] The first terminal can input the semantic communication information into the local semantic knowledge base model stored in the first terminal, and perform semantic processing on the semantic communication information by using the local semantic knowledge base model.

[0080] As described above, in the case where the first terminal is a sending terminal, the local semantic knowledge base model can perform semantic extraction on the semantic communication information to obtain semantic information corresponding to the semantic communication information. The semantic information can be a semantic feature extracted from the semantic communication information. In the case where the first terminal is a receiving terminal, the local semantic knowledge base model can perform semantic restoration on the semantic communication information to obtain semantic information corresponding to the semantic communication information. The semantic information can be complete communication information restored according to the semantic feature.

[0081] The local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0082] Federated machine learning is a multi-party joint modeling through exchanging locally trained model parameters. In the embodiment of the present application, the server can jointly perform federated learning with the first terminal and / or the second terminal, and obtain a global model after learning. The server can distribute the model parameters of the global model to the first terminal and the second terminal. Optionally, the first terminal stores the global model as a local semantic knowledge base model. Optionally, the first terminal can also optimize the global model using local data to obtain a local semantic knowledge base model.

[0083] After the first terminal obtains the semantic communication information in the semantic communication process, the first terminal can perform semantic processing on the semantic communication information through the local semantic knowledge base model, i.e., performing semantic extraction or semantic restoration using the local semantic knowledge base model to obtain semantic information.

[0084] Please continue to refer to FIG. 3. As described above, if the semantic processing fails, the first terminal sends the semantic communication information to the server.

[0085] For example, after the first terminal inputs the semantic communication information into the local semantic knowledge base model, the first terminal does not obtain the semantic information corresponding to the semantic communication information output by the local semantic knowledge base model, and determines that the semantic processing fails. The first terminal sends the semantic communication information to the server. After the server receives the semantic communication information sent by the first terminal, the server performs semantic processing on the semantic communication information, and then sends the semantic information obtained by the semantic processing to the first terminal.

[0086] If the semantic processing is successful, the first terminal obtains the semantic information corresponding to the semantic communication information.

[0087] The above embodiment constructs a local semantic knowledge base model through federated learning. Each party participating in joint modeling optimizes the model in the process of exchanging model parameters, and obtains better model performance. Federated learning can meet the joint modeling of original data not leaving the local and more complex scenarios, and avoid security risks such as data / model privacy leakage in the joint modeling process.

[0088] In one embodiment, based on the embodiment shown in FIG. 2 and in combination with the embodiment shown in FIG. 4, in the present embodiment, the semantic communication method further includes step S401 shown in FIG. 4.

[0089] Step S401: The first terminal receives the semantic information distributed by the server.

[0090] If the server successfully performs semantic processing on the semantic communication information sent by the first terminal, obtains semantic information, the server sends the semantic information to the first terminal, so that the first terminal can receive the semantic information sent by the server.

[0091] The semantic information is obtained by the server performing semantic processing on the semantic communication information according to the global model, or the semantic information is obtained by the server performing semantic processing on the approximate features of the semantic communication information according to the global model.

[0092] In the embodiments of the application, the global model is a semantic knowledge base base model constructed by the server through federated learning. As described above, the server can perform federated learning jointly with the first terminal and / or the second terminal, and obtain the global model after learning. The server can distribute the model parameters of the global model to the first terminal and the second terminal. The first terminal stores the global model as a local semantic knowledge base model. Optionally, the first terminal can also optimize the global model using local data to obtain a local semantic knowledge base model.

[0093] Therefore, after receiving the semantic communication information sent by the first terminal, the server first performs semantic processing on the semantic communication information according to the locally stored global model. For example, the server can input the semantic communication information into the global model and attempt to perform semantic processing on the semantic communication information through the global model to obtain semantic information.

[0094] It can be understood that, similar to the role of the local semantic knowledge base model in the first terminal, in the case of the first terminal as a sending terminal, the locally stored global model in the server can perform semantic extraction to obtain semantic information, which can be semantic features corresponding to the semantic communication information. In the case of the first terminal as a receiving terminal, the locally stored global model in the server can perform semantic restoration to obtain semantic information, which can be complete communication information restored from the semantic features. The following embodiments are described by taking the first terminal as a receiving terminal as an example unless otherwise specified.

[0095] If the server performs semantic processing on the semantic communication information according to the global model and successfully obtains semantic information, the server returns the obtained semantic information to the first terminal.

[0096] If the server performs semantic processing on the semantic communication information according to the global model and does not obtain semantic information, the server performs information disassembly on the semantic communication information to obtain approximate features, and performs semantic processing on the approximate features according to the global model.

[0097] As an implementation, the server can input the semantic communication information into the approximate feature disassembling model to obtain an approximate feature of the semantic communication information; then, the server inputs the approximate feature into the global model, and attempts to obtain semantic information by performing semantic processing on the approximate feature by using the global model; if the semantic information is successfully obtained, the server returns the obtained semantic information to the first terminal.

[0098] In some embodiments, if the semantic information is not obtained by the server in the process of performing semantic processing on the approximate feature by using the global model, it indicates that the semantic processing of the server fails, and the server sends a semantic processing failure feedback to the first terminal. In some embodiments, if the approximate feature is not obtained by the server in the process of disassembling the semantic communication information, it also indicates that the semantic processing of the server fails, and the server sends a semantic processing failure feedback to the first terminal.

[0099] After receiving the semantic processing failure feedback sent by the server, the first terminal sends the semantic processing failure feedback to the second terminal; in some embodiments, the second terminal sends the semantic processing failure feedback to the server, and the server determines to perform semantic processing again (for example, performs information disassembling again to obtain an approximate feature, and performs semantic processing on the disassembled approximate feature by using the global model) to obtain semantic information. The process of performing semantic processing again will be described in the following embodiments. If the server receives a plurality of failure notifications sent by the second terminal, the server deletes the disassembled approximate feature, and ends the current semantic communication process.

[0100] In this way, when the first terminal performs semantic processing on the semantic communication information, if the semantic processing of the first terminal succeeds, the first terminal obtains the semantic information corresponding to the semantic communication information; if the semantic processing of the first terminal fails, the first terminal sends the semantic communication information to the server, and if the semantic processing of the server succeeds, the server returns the semantic information corresponding to the semantic communication information to the first terminal.

[0101] It can be understood that if the semantic processing of the first terminal succeeds to obtain the semantic information, or the semantic processing of the server succeeds to obtain the semantic information, but the semantic information is inaccurate, in this case, it also cannot be determined that the semantic communication process is completed.

[0102] Therefore, referring to FIG. 5, the semantic communication method of the embodiments of the present application further includes steps S501-S504 shown in FIG. 5.

[0103] Step S501: The first terminal determines whether the current semantic communication process is completed according to the semantic information.

[0104] The first terminal determines whether the current semantic communication process is completed according to the semantic information.

[0105] In a possible implementation, when the second terminal sends the semantic communication information to the first terminal, the second terminal can also send the communication target and / or the communication intention corresponding to the current semantic communication process; as the sending terminal, the second terminal perceives the communication environment, obtains the communication information to be transmitted, searches the knowledge base according to the communication information to be transmitted, extracts the semantic communication information (and the communication target and / or the communication intention in some embodiments) to be transmitted, and sends the semantic communication information to the first terminal; the first terminal performs semantic processing according to the received semantic communication information (and the communication target and / or the communication intention sent by the second terminal in some embodiments), or performs semantic processing through the server, to obtain the semantic information. In this way, when the first terminal determines whether the current semantic communication process is completed according to the semantic information, the first terminal can specifically determine whether the semantic information meets the communication intention or the communication target corresponding to the semantic communication information. For example, the first terminal extracts the communication intention or the communication target from the semantic communication information, and then compares the extracted communication intention or the communication target with the communication target or the communication intention sent by the second terminal, to determine whether the semantic information meets the communication intention or the communication target sent by the second terminal. If the semantic information meets the communication intention or the communication target, the first terminal determines that the current semantic communication process is completed, that is, the first terminal determines that the semantic information can form a complete semantic corresponding to the communication intention or the communication target, and then determines that the current semantic communication process is completed. In this way, the communication parties can reach an agreement on the judgment standard of whether the communication is completed, and normal semantic communication can be ensured.

[0106] In another possible implementation, the first terminal can also input the semantic communication information and the corresponding semantic information into a pre-trained classification model, to determine whether the semantic information meets the communication intention or the communication target corresponding to the semantic communication information.

[0107] Step S502: If it is determined that the current semantic communication process is completed, the first terminal sends semantic communication completion information to the second terminal.

[0108] Step S503: The first terminal performs an operation corresponding to the semantic information.

[0109] If it is determined that the current semantic communication process is completed, the first terminal sends semantic communication completion information to the second terminal. In some embodiments, if the semantic information indicates that the first terminal performs an operation, the first terminal performs the operation corresponding to the semantic information.

[0110] Step S504: If it is determined that the current semantic communication process is not completed, the first terminal sends feedback information to the second terminal.

[0111] Specifically, the feedback information is used to instruct the second terminal to re-perform the semantic communication process.

[0112] The second terminal judges the context information according to the feedback information of the first terminal, re-obtains the semantic communication information, for example, the second terminal re-performs semantic extraction on the communication information to be transmitted, so that the semantic features can be extracted from the communication information to be transmitted as the semantic communication information that needs to be re-sent to the first terminal, and the second terminal sends the semantic communication information to the first terminal.

[0113] After the first terminal receives the semantic communication information re-sent by the second terminal, the first terminal performs semantic processing on the semantic communication information re-sent by the second terminal to obtain semantic information.

[0114] Similar to the embodiment shown in FIG. 2, the first terminal first attempts to perform semantic processing on the semantic communication information in the semantic communication process by itself, and if the semantic processing is successful, the first terminal obtains semantic information; if the semantic processing fails, the first terminal re-sends the semantic communication information to the server for re-performing semantic processing on the semantic communication information; the server first performs semantic processing on the semantic communication information according to the locally stored global model, and if the semantic information is successfully obtained, the server returns the obtained semantic information to the first terminal; if the server does not obtain the semantic information after performing semantic processing on the semantic communication information according to the global model, the server performs information disassembly on the semantic communication information to obtain approximate features, and performs semantic processing on the approximate features according to the global model, and if the semantic information is successfully obtained, the server returns the obtained semantic information to the first terminal, and if the semantic processing fails, the server sends a semantic processing failure feedback to the first terminal.

[0115] In this embodiment, when the first terminal and the server perform semantic processing, the context information can also be combined to search the local semantic knowledge base basic model to analyze the semantic communication information to obtain the semantic information.

[0116] In the above embodiment, the server performs information decomposition on the semantic communication information sent by the first terminal to obtain the approximate feature. When the server successfully obtains the semantic information by performing semantic processing on the approximate feature through the global model, the server returns the obtained semantic information to the first terminal. When the first terminal determines that the current semantic communication process is completed according to the semantic information, the first terminal sends semantic communication completion information to the second terminal.

[0117] After the server receives the semantic communication completion information sent by the second terminal, the server updates the model parameters of the global model according to the approximate feature to obtain updated model parameters, that is, the server trains the global model by taking the approximate feature as a training sample to optimize the global model, and obtains the latest model parameters of the global model after optimization: After updating the model parameters, the server sends the updated model parameters to the first terminal for updating the local semantic knowledge base model; in some embodiments, the server can also perform corresponding operations according to the indication of the semantic communication completion information.

[0118] After the first terminal receives the updated model parameters sent by the server, the first terminal updates the local semantic knowledge base model based on the updated model parameters.

[0119] In this way, the server and the terminal in the above embodiment can update the semantic knowledge base basic model according to the processing capability and communication data of the server and the terminal, maximize the coverage of various approximate features, thereby improving the accuracy of communication and ensuring the quality of communication.

[0120] In one embodiment, as shown in FIG. 6, the present application provides a semantic communication method for a server, which includes the following steps S601-S602.

[0121] Step S601: The server receives semantic communication information sent by the first terminal and performs semantic processing on the semantic communication information.

[0122] Step S602: The server sends the semantic information obtained by the semantic processing to the first terminal.

[0123] In one embodiment, based on the embodiment shown in FIG. 6 and in combination with the embodiment shown in FIG. 7, step S601 includes steps S701, S702 and S703 shown in FIG. 7.

[0124] Step S701: The server receives semantic communication information sent by the first terminal.

[0125] Step S702: The server performs semantic processing on the semantic communication information according to the global model.

[0126] Step S703: If the semantic information is not obtained, the server performs information decomposition on the semantic communication information to obtain approximate features, and performs semantic processing on the approximate features through the global model.

[0127] In one embodiment, the method further comprises:

[0128] The server receives semantic communication completion information sent by the second terminal, wherein the semantic communication completion information is determined by the first terminal according to the semantic information, and the semantic information is obtained by the server through semantic processing on the approximate features through the global model.

[0129] The server updates the model parameters of the global model according to the approximate features, to obtain updated model parameters.

[0130] The server sends the updated model parameters to the first terminal.

[0131] In one embodiment, the method further comprises:

[0132] If the semantic information is not obtained in the process of performing semantic processing on the approximate features through the global model, the server sends a semantic processing failure feedback to the first terminal.

[0133] In one embodiment, the method further comprises:

[0134] If the server receives multiple failure notifications sent by the second terminal, the server deletes the approximate features obtained through the decomposition.

[0135] For the implementation of the semantic communication method for the server, please refer to the related description in the implementation of the semantic communication method for the first terminal in the foregoing. In addition, the beneficial effects of the semantic communication method for the server are similar to those of the semantic communication method for the first terminal in the foregoing, and will not be repeated here.

[0136] In one embodiment of the present application, a semantic communication method is provided, as shown in FIG. 8, comprising the following steps S801-S808.

[0137] Step S801: The first terminal receives semantic communication information sent by the second terminal.

[0138] Step S802: The first terminal performs semantic processing on the semantic communication information through the local semantic knowledge base model, and if the semantic processing fails, the first terminal sends the semantic communication information to the server.

[0139] Step S803: The server performs semantic processing on the semantic communication information through the global model, and if the semantic information is not obtained, the server performs information decomposition on the semantic communication information to obtain approximate features, and performs semantic processing on the approximate features through the global model.

[0140] Step S804: The server sends the semantic information obtained by the semantic processing to the first terminal.

[0141] Step S805: The first terminal determines whether the current semantic communication process is completed according to the semantic information, and if the current semantic communication process is completed, the first terminal sends semantic communication completion information to the second terminal.

[0142] Step S806: The second terminal sends the semantic communication completion information to the server.

[0143] Step S807: After the server receives the semantic communication completion information sent by the second terminal, the server updates the model parameters of the global model according to the approximate feature, and obtains updated model parameters.

[0144] Step S808: The server sends the updated model parameters to the first terminal and / or the second terminal.

[0145] In an embodiment of the present application, a semantic communication method is provided, as shown in FIG. 9, including the following steps S901 to S912.

[0146] Step S901: The first terminal receives the semantic communication information sent by the second terminal.

[0147] Step S902: The first terminal performs semantic processing on the semantic communication information by using the local semantic knowledge base model, and if the semantic processing fails, the first terminal sends the semantic communication information to the server.

[0148] Step S903: The server performs semantic processing on the semantic communication information by using the global model, and if the semantic information is not obtained, the server performs information disassembly on the semantic communication information to obtain an approximate feature, and performs semantic processing on the approximate feature by using the global model.

[0149] Step S904: The server sends the semantic information obtained by the semantic processing to the first terminal.

[0150] Step S905: The first terminal determines whether the current semantic communication process is completed according to the semantic information, and if the current semantic communication process is not completed, the first terminal sends feedback information to the second terminal.

[0151] Step S906: The second terminal re-sends the semantic communication information to the first terminal.

[0152] Step S907: The second terminal sends a semantic processing failure feedback to the server.

[0153] Step S908: The first terminal performs semantic processing on the semantic communication information re-sent by the second terminal, and if the semantic processing fails, the first terminal re-sends the semantic communication information to the server.

[0154] Step S909: The server re-performs semantic processing on the semantic communication information through the global model. If semantic information is not obtained, the server re-performs information disassembly on the semantic communication information to obtain approximate features, and performs semantic processing on the approximate features through the global model. The server sends the semantic information obtained through the semantic processing to the first terminal.

[0155] Step S910: The first terminal determines whether the current semantic communication process is completed according to the semantic information. If the current semantic communication process is not completed, the first terminal sends feedback information to the second terminal.

[0156] Step S911: If the number of times that the semantic communication process is not completed reaches a threshold, the second terminal sends a multiple failure notification to the server.

[0157] Step S912: After receiving the multiple failure notification, the server deletes the approximate features obtained through the disassembly, and retains the model parameters of the global model.

[0158] In an embodiment of the present application, a semantic communication method is provided, as shown in FIG. 10, including the following steps S1001 to S1004.

[0159] Step S1001: The first terminal receives semantic communication information sent by the second terminal.

[0160] Step S1002: The first terminal performs semantic processing on the semantic communication information by using a local semantic knowledge base model. If the semantic processing is successful, the first terminal obtains semantic information.

[0161] Step S1003: The first terminal determines whether the current semantic communication process is completed according to the semantic information.

[0162] Step S1004: If the current semantic communication process is completed, the first terminal sends semantic communication completion information to the second terminal.

[0163] In the following, two examples are combined to illustrate the semantic communication method of the present application.

[0164] Example 1: There are multiple smart home terminals: terminal 1, terminal 2, terminal 3 and terminal 4 (in the case of the above-mentioned first terminal as a receiving end and the second terminal as a sending end, terminal 1 is equivalent to the above-mentioned second terminal, and terminal 2, terminal 3 and terminal 4 are equivalent to the above-mentioned first terminal), wherein terminal 1 is a mobile phone, terminal 2 is a computer, terminal 3 is a sweeping robot, and terminal 4 is an air conditioner. The multiple smart home terminals can be connected to an indoor home computing center (equivalent to the above-mentioned server).

[0165] 1) Terminal 1 senses the communication environment and obtains the communication scenario information: Terminal 1 is in an outdoor environment, the current time is the preparation for leaving work time, the environment noise interference of terminal 1 is small, the available transmission bandwidth resource is large, and terminal 1 needs to set the indoor terminal to the welcome home mode.

[0166] 2) Terminal 1 determines the communication target according to the sensed communication scenario information: the indoor terminal needs to be set to the welcome home mode, and then terminal 1 searches the knowledge base according to the communication scenario information and extracts the semantic communication information (optionally, the communication intention can also be extracted) that needs to be transmitted. Terminal 1 sends the semantic communication information (optionally, carrying the communication target and / or the communication intention) to terminal 2, terminal 3 and terminal 4 respectively.

[0167] 3) Terminal 2 senses the communication environment: terminal 2 is in a regular position, it is 5:00 now, and it is in standby state; terminal 3 senses the communication environment: terminal 3 is in the bedroom, it is 5:00 now, and it is in the state of sweeping and mopping; terminal 4 senses the communication environment: terminal 4 is in a regular position, it is 5:00 now, the indoor temperature is 30°, and the humidity is 90%.

[0168] 4) Terminal 2, terminal 3 and terminal 4 receive the semantic communication information from terminal 1, and terminal 2, terminal 3 and terminal 4 respectively perform semantic processing (semantic restoration) on the semantic communication information to try to restore the complete semantic information corresponding to the semantic communication information.

[0169] 5) Terminal 2 obtains the semantic information after semantic processing: set to welcome mode, terminal 2 judges that the semantic information is complete and conforms to the communication intention, and terminal 2 determines that the current semantic communication process is completed.

[0170] Terminal 2 switches from standby mode to on mode and sends the semantic communication completion information to terminal 1, ending the communication.

[0171] 6) Terminal 3 and terminal 4 fail in semantic processing, and terminal 3 and terminal 4 respectively transmit the received semantic communication information to the indoor home computing center.

[0172] 7) The indoor home computing center performs semantic processing (semantic restoration) on the semantic communication information through the global model.

[0173] 8) The indoor home computing center successfully performs semantic processing and obtains the semantic information: set to welcome home mode, and the indoor home computing center sends the semantic information to terminal 3 and terminal 4.

[0174] 9) Terminal 3 judges that the semantic information is complete and conforms to the communication intention, and terminal 3 determines that the current semantic communication process is completed.

[0175] Terminal 3 switches from working mode to standby mode and sends semantic communication completion information to terminal 1, ending the communication.

[0176] 10) Terminal 4 judges that the semantic information is complete and meets the communication intention, and terminal 4 determines that the current semantic communication process is complete.

[0177] Terminal 4 turns on the air conditioner and sets the temperature to 23°, and sends semantic communication completion information to terminal 1, ending the communication.

[0178] Example two: There are multiple meta-universe terminals: terminal 1, terminal 2, terminal 3 and terminal 4 (in the case of the above first terminal as the receiving end and the second terminal as the sending end, terminal 1 is equivalent to the above-mentioned second terminal, and terminal 2, terminal 3 and terminal 4 are equivalent to the above-mentioned first terminal), wherein terminal 1 is a game host, terminal 2 is an extended reality (Extended Reality, ER) glasses, and terminal 2 is connected to a computing power server 1; terminal 3 is an augmented reality (Augmented Reality, AR) glasses, and terminal 3 is connected to a computing power server 1; terminal 4 is a normal translation glasses, and terminal 4 is connected to a computing power server 2.

[0179] 1) Terminal 1 perceives the communication environment and obtains communication scene information: terminal 1 has more available transmission bandwidth resources.

[0180] 2) Terminal 1 determines the communication target according to the perceived communication scene information: it is necessary to update the game content, and then terminal 1 searches the knowledge base according to the communication scene information and extracts the semantic communication information (optionally, the communication intention can also be extracted) that needs to be transmitted, and terminal 1 sends the semantic communication information (optionally, carrying the communication target and / or the communication intention) to terminal 2, terminal 3 and terminal 4 respectively.

[0181] 3) Terminal 2 perceives the communication environment: terminal 2 is in a regular position, has connected a wireless fidelity (Wireless Fidelity, WIFI) network, and is in a wearing state; terminal 3 perceives the communication environment: terminal 3 is in a bedroom, has connected a WIFI, and is in a meeting state; terminal 4 perceives the communication environment: terminal 4 is in a regular position, and is in a data mode.

[0182] 4) Terminal 2, terminal 3 and terminal 4 receive the semantic communication information from terminal 1, and terminal 2, terminal 3 and terminal 4 respectively perform semantic processing (semantic restoration) on the semantic communication information to try to restore the complete semantic information corresponding to the semantic communication information.

[0183] 5) Terminal 2 obtains the semantic information after semantic processing: skin update, terminal 2 judges that the semantic information is complete and meets the communication intention, and terminal 2 determines that the current semantic communication process is complete.

[0184] Terminal 2 updates the game skin and sends semantic communication completion information to terminal 1, ending the communication.

[0185] 6) Terminal 3 and terminal 4 fail in semantic processing, terminal 3 transmits the received semantic communication information to computing power server 1, and terminal 4 transmits the received semantic communication information to computing power server 2.

[0186] 7) Computing power server 1 performs semantic processing (semantic restoration) on the semantic communication information through the global model, and obtains semantic information: update skin state. Computing power server 1 issues the semantic information to terminal 3.

[0187] 8) Terminal 3 judges that the semantic information is complete in semantics and conforms to the communication intention, and terminal 3 determines that the current semantic communication process is completed, and performs game skin update; if the semantic information does not conform to the communication intention or the communication target, the game skin update is not performed.

[0188] Terminal 3 sends semantic communication completion information to terminal 1, ending the communication.

[0189] 9) Computing power server 2 fails to successfully perform semantic processing (semantic restoration) on the semantic communication information through the global model.

[0190] 10) Computing power server 2 performs information disassembly on the semantic communication information to obtain approximate features: lens color (at this time, the approximate features disassembled by computing power server 2 are inaccurate), and computing power server 2 inputs the approximate features into the global model to obtain extraction results (semantic information) returned to terminal 4.

[0191] 11) Terminal 4 determines that the current semantic communication process is not completed according to the semantic information, and sends feedback information to terminal 1 to re-perform communication.

[0192] 12) Terminal 1 judges the context information according to the feedback information of terminal 4, re-searches the knowledge base, and re-obtains semantic communication information: update game character skin, and sends to terminal 4.

[0193] 13) Terminal 4 re-performs semantic processing (semantic restoration) according to the received semantic communication information, and fails to restore the semantic information. Terminal 4 sends the semantic communication information to computing power server 2.

[0194] 14) Computing power server 2 performs information disassembly on the semantic communication information, fails to extract approximate features, and sends semantic processing failure feedback to terminal 4. Computing power server 2 deletes related data and retains the original model parameters without updating.

[0195] 15) Terminal 4 feeds back to terminal 1: semantic processing failure feedback.

[0196] It should be understood that although the steps in the above flowcharts are shown in sequence according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other orders. Moreover, at least part of the steps in the above flowcharts can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of the steps or stages is not necessarily sequential, but can be alternately or alternately executed with at least part of other steps or steps or stages in other steps.

[0197] In an embodiment of the present application, as shown in FIG. 11, a semantic communication device is also provided, arranged in a first terminal, comprising:

[0198] The semantic processing module 1101 is configured to perform semantic processing on the semantic communication information in the semantic communication process.

[0199] The sending module 1102 is configured to send the semantic communication information to a server if the semantic processing fails, so that the server processes the semantic communication information and obtains semantic information corresponding to the semantic communication information, and returns the semantic information.

[0200] In an embodiment, the semantic processing includes semantic extraction and / or semantic restoration.

[0201] In an embodiment, the semantic processing module 1101 is specifically configured to perform semantic processing on the semantic communication information by using a local semantic knowledge base model.

[0202] In an embodiment, the local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0203] In an embodiment, the above device further comprises:

[0204] The receiving module is configured to receive semantic communication information sent by a second terminal.

[0205] In an embodiment, the receiving module is further configured to receive semantic information issued by a server.

[0206] In an embodiment, the semantic information is obtained by the server by performing semantic processing on the semantic communication information by using a global model, or the semantic information is obtained by the server by performing semantic processing on the approximate features of the semantic communication information by using the global model.

[0207] In an embodiment, the above device further comprises:

[0208] The obtaining module is configured to obtain the semantic information if the semantic processing is successful.

[0209] In an embodiment, the apparatus further includes:

[0210] The determining module is configured to determine whether the current semantic communication process is completed according to the semantic information.

[0211] The sending module 1102 is further configured to send semantic communication completion information to the second terminal if the current semantic communication process is completed.

[0212] In an embodiment, the apparatus further includes:

[0213] The executing module is configured to execute an operation corresponding to the semantic information.

[0214] In an embodiment, the determining module is specifically configured to determine whether the semantic information meets a communication intention or a communication target corresponding to the semantic communication information; and determine that the current semantic communication process is completed if the semantic information meets the communication intention or the communication target.

[0215] In an embodiment, the sending module 1102 is further configured to send feedback information to the second terminal if the current semantic communication process is not completed, where the feedback information is used to instruct the second terminal to re-execute the semantic communication process.

[0216] In an embodiment, the receiving module is further configured to receive semantic communication information re-sent by the second terminal, and perform semantic processing on the semantic communication information re-sent by the second terminal to obtain the semantic information.

[0217] In an embodiment, the receiving module is further configured to receive updated model parameters sent by the server, where the updated model parameters are obtained by performing model parameter updating on the global model according to the approximate feature after the server performs semantic processing on the approximate feature to obtain the semantic information, and receives the semantic communication completion information sent by the second terminal.

[0218] The apparatus further includes:

[0219] The updating module is configured to update the local semantic knowledge base model based on the updated model parameters.

[0220] In an embodiment, the receiving module is further configured to receive semantic processing failure feedback sent by the server.

[0221] The sending module 1102 is further configured to send the semantic processing failure feedback to the second terminal.

[0222] For specific limitations of the semantic communication device arranged in the first terminal, reference can be made to the limitations of the semantic communication method used in the first terminal as described above, which will not be repeated here. Each module in the above semantic communication device can be realized by software, hardware and their combination in whole or in part. The above modules can be embedded in or independent of the processor in the first terminal in hardware form, or stored in the memory in the first terminal in software form, so as to be called and executed by the processor to perform the operations corresponding to each of the above modules.

[0223] In an embodiment of the present application, as shown in FIG. 12, a semantic communication device arranged in the second terminal is also provided, which comprises:

[0224] The receiving module 1201 is configured to receive the semantic communication information sent by the first terminal and perform semantic processing on the semantic communication information.

[0225] The sending module 1202 is configured to send the semantic information obtained by the semantic processing to the first terminal.

[0226] In an embodiment, the receiving module 1201 is specifically configured to perform semantic processing on the semantic communication information by using the global model; if no semantic information is obtained, the semantic communication information is disassembled to obtain approximate features, and the approximate features are processed by the global model.

[0227] In an embodiment, the receiving module 1201 is further configured to receive the semantic communication completion information sent by the second terminal, wherein the semantic communication completion information is determined by the first terminal according to the semantic information obtained by the server by processing the approximate features by using the global model.

[0228] The above device further comprises:

[0229] The updating module is configured to update the model parameters of the global model according to the approximate features to obtain updated model parameters.

[0230] The sending module 1202 is further configured to send the updated model parameters to the first terminal.

[0231] In an embodiment, the sending module 1202 is further configured to send a semantic processing failure feedback to the first terminal if no semantic information is obtained in the process of processing the approximate features by using the global model.

[0232] In an embodiment, the above device further comprises:

[0233] The deleting module is configured to delete the disassembled approximate features if the multiple failure notifications sent by the second terminal are received.

[0234] The specific limitation of the semantic communication device arranged in the server can refer to the limitation of the semantic communication method used in the server as described above, which will not be repeated here. Each module in the semantic communication device described above can be realized by software, hardware and their combination in whole or in part. The above-mentioned modules can be embedded in or independent of the processor in the server in hardware form, or can be stored in the memory in the server in software form, so as to be called and executed by the processor to perform the operation corresponding to each module.

[0235] FIG. 13 is a structural schematic diagram of a terminal provided by an embodiment of the present application. The terminal 1300 shown in FIG. 13 includes at least one processor 1301, a memory 1302, at least one network interface 1304 and a user interface 1303. Each component in the terminal 1300 is coupled together through a bus system 1305. It can be understood that the bus system 1305 is used to realize the connection communication between the components. In addition to the data bus, the bus system 1305 also includes a power bus, a control bus and a status signal bus. However, in order to clearly illustrate, all kinds of buses are marked as the bus system 1305 in FIG. 1. In addition, the terminal 1300 described in the embodiment of the present application further includes a transceiver 1306, which can be multiple elements, that is, includes a transmitter and a receiver, and provides a unit for communicating with various other devices on a transmission medium.

[0236] The user interface 1303 can include a display, a keyboard or a clicking device (for example, a mouse, a trackball, a touchpad or a touch screen, etc.).

[0237] It is to be appreciated that the memory 1302 in the embodiments of the present application can be a volatile memory or a nonvolatile memory, or can include both volatile and nonvolatile memory. Among them, the nonvolatile memory can be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable Programmable ROM (EPROM), an Electrically EPROM (EEPROM), or a flash memory. The volatile memory can be a Random Access Memory (RAM) used as an external cache. By way of example, and not limitation, many forms of RAM can be used, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 1302 of the system and method described in the embodiments of the present application is intended to include, without being limited to, these and any other suitable types of memory.

[0238] In some embodiments, the memory 1302 stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof: an operating system 13021 and an application program 13022.

[0239] Among them, the operating system 13021 contains various system programs, such as framework layer, core library layer, driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application program 13022 contains various application programs, such as Media Player, Browser, etc., for implementing various application services. The program for implementing the method of the embodiments of the present application can be contained in the application program 13022.

[0240] In the embodiments of the present application, the program or instruction stored in the memory 1302 is called, specifically, the program or instruction stored in the application program 13022. Among them, the processor 1301 is used to perform semantic processing on the semantic communication information in the semantic communication process.

[0241] The transmitter is configured to send the semantic communication information to the server if the semantic processing fails, so as to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and return the semantic information.

[0242] The processor 1301 can be an integrated circuit chip having a signal processing capability. In implementation process, the steps of the above method can be completed by an integrated logic circuit or a software form of instruction in the processor 1301. The processor 1301 can 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 devices, discrete gate or transistor logic devices, discrete hardware components. The methods, steps and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed by the processor 1301. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the methods disclosed in conjunction with the embodiments of the present application can be directly embodied as a hardware code executed by the processor, or a combination of hardware and software modules in the processor. The software module can reside in a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, or any other forms of storage medium in the art. The storage medium is located in the storage 1302, and the processor 1301 reads information in the storage 1302 and combines the hardware to complete the steps of the above method.

[0243] It can be understood that the embodiments described in the present application can be realized by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field-Programmable Gate Arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units for executing the functions described in the present application or a combination thereof.

[0244] For software implementation, the techniques described in the embodiments of the present application can be implemented by means of modules (for example, procedures, functions, and so on) that perform the functions described in the embodiments of the present application. Software codes can be stored in the memory and executed by the processor 1301. The memory can be implemented in the processor 1301 or outside the processor 1301.

[0245] In one embodiment, the semantic processing includes semantic extraction and / or semantic restoration.

[0246] In one embodiment, the processor 1301 is specifically configured to perform semantic processing on the semantic communication information by using the local semantic knowledge base model.

[0247] In one embodiment, the local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0248] In one embodiment, the receiver is configured to receive the semantic communication information sent by the second terminal.

[0249] In one embodiment, the receiver is further configured to receive the semantic information issued by the server.

[0250] The semantic information is obtained by the server by performing semantic processing on the semantic communication information through a global model, or the semantic information is obtained by the server by performing semantic processing on the approximate features of the semantic communication information through the global model.

[0251] In one embodiment, the processor 1301 is further configured to obtain the semantic information if the semantic processing is successful.

[0252] In one embodiment, the processor 1301 is further configured to determine whether the current semantic communication process is completed according to the semantic information.

[0253] The transmitter is further configured to send the semantic communication completion information to the second terminal if the current semantic communication process is completed.

[0254] In one embodiment, the processor 1301 is further configured to perform an operation corresponding to the semantic information.

[0255] In one embodiment, the processor 1301 is specifically configured to determine whether the semantic information meets a communication intention or a communication target corresponding to the semantic communication information, and determine that the current semantic communication process is completed if the semantic information meets the communication intention or the communication target.

[0256] In one embodiment, the transmitter is further configured to send feedback information to the second terminal if the current semantic communication process is not completed, the feedback information being used to instruct the second terminal to re-execute the semantic communication process.

[0257] In an embodiment, the receiver is further configured to receive the semantic communication information retransmitted by the second terminal, and perform semantic processing on the semantic communication information retransmitted by the second terminal to obtain the semantic information.

[0258] In an embodiment, the receiver is further configured to receive updated model parameters transmitted by the server, the updated model parameters being obtained by performing model parameter updating on the global model according to the approximate feature after the server performs semantic processing on the approximate feature to obtain the semantic information, and after receiving the semantic communication completion information transmitted by the second terminal;

[0259] The processor is further configured to update the local semantic knowledge base model based on the updated model parameters.

[0260] In an embodiment, the receiver is further configured to receive semantic processing failure feedback transmitted by the server.

[0261] The transmitter is further configured to transmit the semantic processing failure feedback to the second terminal.

[0262] FIG. 14 is a structural schematic diagram of a server provided in an embodiment of the present application. The server 1400 shown in FIG. 14 includes at least one processor 1401, a memory 1402, and at least one network interface 1404. The various components in the server 1400 are coupled together by a bus system 1405. It can be understood that the bus system 1405 is configured to realize the connection and communication between the components. The bus system 1405 includes not only a data bus, but also a power supply bus, a control bus, and a status signal bus. However, for the purpose of clarity, all the buses are marked as the bus system 1405 in FIG. 14. In addition, in an embodiment of the present application, the server further includes a transceiver 1406, which can be multiple elements, i.e., includes a transmitter and a receiver, and provides a unit for communicating with various other devices over a transmission medium.

[0263] It can be understood that the memory 1402 in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (Read-Only Memory, ROM), a programmable read-only memory (Programmable ROM, PROM), an erasable programmable read-only memory (Erasable PROM, EPROM), an electrically erasable programmable read-only memory (Electrically EPROM, EEPROM) or a flash memory. The volatile memory can be a random access memory (Random Access Memory, RAM) used as an external cache. By way of example, but not by way of limitation, many forms of RAM are available, such as static random access memory (Static RAM, SRAM), dynamic random access memory (Dynamic RAM, DRAM), synchronous dynamic random access memory (Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (Double Data Rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (Enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (Synchlink DRAM, SLDRAM) and direct memory bus random access memory (Direct Rambus RAM, DRRAM). The memory 1402 of the system and method described in the embodiments of the present application is intended to include, but not limited to, these and any other suitable types of memory.

[0264] In some embodiments, the memory 1402 stores an operating system 14021, an executable module or a data structure, or a subset thereof, or an extended set thereof. Among them, the operating system 14021 contains various system programs, such as framework layer, core library layer, driver layer, etc., for implementing various basic services and processing hardware-based tasks.

[0265] In the embodiments of the present application, by calling the programs or instructions stored in the memory 1402, the receiver is used to receive the semantic communication information sent by the first terminal; the processor 1401 is used to perform semantic processing on the semantic communication information; and the transmitter is used to send the semantic information obtained by the semantic processing to the first terminal.

[0266] Part or all of the methods disclosed in the embodiments of the present application can also be applied in the processor 1401, or implemented by the processor 1401, or implemented by the processor 1401 in cooperation with other elements (for example, a transceiver). The processor 1401 can be an integrated circuit chip having a processing capability. In the implementation process, each step of the above method can be completed by integrated logic circuits or instructions in the form of software in the processor 1401. The above processor 1401 can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. Each method, step and logic block diagram disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read only memory, a programmable read only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory 1402, and the processor 1401 reads the information in the memory 1402, and combines the hardware to complete the steps of the above method.

[0267] It can be understood that the embodiments described in the embodiments of the present application can be realized by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be realized in one or more application specific integrated circuits (Application Specific Integrated Circuits, ASIC), digital signal processors (Digital Signal Processing, DSP), digital signal processing devices (DSP Device, DSPD), programmable logic devices (Programmable Logic Device, PLD), field programmable gate arrays (Field-Programmable Gate Array, FPGA), general processors, controllers, microcontrollers, microprocessors, other electronic units for executing functions described in the present application or a combination thereof.

[0268] For software implementation, the techniques described in the embodiments of the present application can be implemented by means of a module (for example, procedures, functions, and so on) for performing the functions described in the embodiments of the present application. Software codes can be stored in a memory and executed by the processor 1401. The memory can be implemented in the processor 1401 or outside the processor 1401.

[0269] In one embodiment, the processor 1401 is specifically configured to perform semantic processing on the semantic communication information by using the global model; if semantic information is not obtained, the semantic communication information is disassembled to obtain approximate features, and the approximate features are processed by using the global model.

[0270] In one embodiment, the receiver is further configured to receive semantic communication completion information sent by the second terminal, the semantic communication completion information being determined by the first terminal according to semantic information, the semantic information being obtained by the server by performing semantic processing on the approximate features by using the global model.

[0271] The processor 1401 is further configured to perform model parameter updating on the global model according to the approximate features, to obtain updated model parameters.

[0272] The transmitter is further configured to send the updated model parameters to the first terminal.

[0273] In one embodiment, the transmitter is further configured to send a semantic processing failure feedback to the first terminal if semantic information is not obtained in the process of performing semantic processing on the approximate features by using the global model.

[0274] In one embodiment, if the receiver receives a plurality of failure notifications sent by the second terminal, the processor 1401 is further configured to delete the disassembled approximate features.

[0275] In one embodiment of the present application, a semantic communication system is also provided, which can be combined with FIG. 1. The semantic communication system includes a first terminal, a second terminal, and a server.

[0276] The second terminal is configured to send semantic communication information to the first terminal.

[0277] The first terminal is configured to perform the steps of the semantic communication method for the first terminal in the above embodiments.

[0278] The server is configured to perform the steps of the semantic communication method for the server in the above embodiments.

[0279] In one embodiment, the second terminal is further configured to perform knowledge base searching according to the to-be-transmitted communication information, to obtain the semantic communication information.

[0280] The second terminal is further configured to send the to-be-transmitted communication information to the server to perform a knowledge base search according to the to-be-transmitted communication information, to obtain semantic communication information, if the knowledge base search fails.

[0281] For the implementation and advantages of the semantic communication system, refer to the relevant description of the above embodiments, which will not be repeated here.

[0282] In an embodiment of the present application, a computer readable storage medium is also provided, which stores a computer program. The computer program is executed by a processor to implement the following steps:

[0283] The semantic communication information in the semantic communication process is subjected to semantic processing.

[0284] If the semantic processing fails, the semantic communication information is sent to the server to process the semantic communication information and obtain semantic information corresponding to the semantic communication information, and return the semantic information.

[0285] In an embodiment, the semantic processing includes semantic extraction and / or semantic restoration.

[0286] In an embodiment, the computer program is executed by the processor to specifically implement the following steps:

[0287] The semantic communication information is subjected to semantic processing by using a local semantic knowledge base model.

[0288] In an embodiment, the local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0289] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0290] The semantic communication information sent by the second terminal is received.

[0291] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0292] The semantic information issued by the server is received.

[0293] The semantic information is obtained by the server through semantic processing of the semantic communication information by using a global model, or the semantic information is obtained by the server through semantic processing of approximate features of the semantic communication information by using the global model.

[0294] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0295] If the semantic processing is successful, the semantic information is obtained.

[0296] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0297] According to the semantic information, it is determined whether the current semantic communication process is completed.

[0298] If the current semantic communication process is completed, semantic communication completion information is sent to the second terminal.

[0299] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0300] An operation corresponding to the semantic information is performed.

[0301] In one embodiment, the computer program, when executed by the processor, specifically implements the following steps:

[0302] It is determined whether the semantic information meets a communication intention or a communication target corresponding to the semantic communication information.

[0303] If the semantic information meets the communication intention or the communication target, it is determined that the current semantic communication process is completed.

[0304] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0305] If the current semantic communication process is not completed, feedback information is sent to the second terminal, and the feedback information is used to instruct the second terminal to re-execute the semantic communication process.

[0306] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0307] The semantic communication information re-sent by the second terminal is received, and the semantic communication information re-sent by the second terminal is semantically processed to obtain the semantic information.

[0308] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0309] The updated model parameter sent by the server is received, and the updated model parameter is obtained after the server semantically processes the approximate feature to obtain the semantic information, and then receives the semantic communication completion information sent by the second terminal, and then updates the model parameter of the global model according to the approximate feature.

[0310] Based on the updated model parameter, the local semantic knowledge base model is updated.

[0311] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0312] The semantic processing failure feedback sent by the server is received.

[0313] send the semantic processing failure feedback to the second terminal.

[0314] In an embodiment of the present application, a computer readable storage medium is also provided, which stores a computer program. The computer program is executed by a processor to implement the following steps:

[0315] receive the semantic communication information sent by the first terminal, and perform semantic processing on the semantic communication information;

[0316] send the semantic information obtained by the semantic processing to the first terminal.

[0317] In an embodiment, the computer program is executed by the processor to specifically implement the following steps:

[0318] perform semantic processing on the semantic communication information by using the global model;

[0319] if the semantic information is not obtained, perform semantic processing on the approximate feature obtained by the information disassembly by using the global model.

[0320] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0321] receive the semantic communication completion information sent by the second terminal, wherein the semantic communication completion information is determined by the first terminal according to the semantic information, and the semantic information is obtained by performing semantic processing on the approximate feature by using the global model by the server;

[0322] update the model parameters of the global model according to the approximate feature, to obtain updated model parameters;

[0323] send the updated model parameters to the first terminal.

[0324] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0325] if the semantic information is not obtained by performing semantic processing on the approximate feature by using the global model, send the semantic processing failure feedback to the first terminal.

[0326] In an embodiment, the computer program is executed by the processor to further implement the following steps:

[0327] if the multiple failure notifications sent by the second terminal are received, delete the approximate feature obtained by the disassembly.

[0328] FIG. 15 is a schematic structural diagram of a chip according to an embodiment of the present application. The chip 1500 shown in FIG. 15 includes a processor 1510. The processor 1510 can call and run a computer program from a memory to implement the method according to an embodiment of the present application.

[0329] Optionally, as shown in FIG. 15, the chip 1500 can further include a memory 1520. The processor 1510 can invoke and run a computer program from the memory 1520 to implement the method in the embodiments of the present application.

[0330] The memory 1520 can be a separate device independent of the processor 1510, or can be integrated in the processor 1510.

[0331] In some embodiments, the chip 1500 can further include an input interface 1530. The processor 1510 can control the input interface 1530 to communicate with other devices or chips, and specifically, can obtain information or data sent by other devices or chips.

[0332] In some embodiments, the chip 1500 can further include an output interface 1540. The processor 1510 can control the output interface 1540 to communicate with other devices or chips, and specifically, can output information or data to other devices or chips.

[0333] In some embodiments, the chip 1500 can be applied to a network function or a base station in the embodiments of the present application, and the chip 1500 can implement the corresponding processes implemented in the methods of the embodiments of the present application. For brevity, details are not repeated here.

[0334] It should be understood that the chip 1500 mentioned in the embodiments of the present application can also be referred to as a system-level chip, a system chip, a chip system, or a system-on-chip, etc.

[0335] The embodiments of the present application also provide a computer program product containing instructions, which are executed by a processor to implement the following steps:

[0336] performing semantic processing on semantic communication information in a semantic communication process;

[0337] If the semantic processing fails, the semantic communication information is sent to a server to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and the semantic information is returned.

[0338] In one embodiment, the semantic processing includes semantic extraction and / or semantic restoration.

[0339] In one embodiment, the computer program is executed by the processor to specifically implement the following steps:

[0340] performing semantic processing on the semantic communication information by using a local semantic knowledge base model.

[0341] In one embodiment, the local semantic knowledge base model is a semantic knowledge base base model constructed by the first terminal through federated learning.

[0342] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0343] receiving semantic communication information sent by a second terminal.

[0344] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0345] receiving semantic information issued by a server;

[0346] The semantic information is obtained by the server performing semantic processing on the semantic communication information through a global model, or the semantic information is obtained by the server performing semantic processing on approximate features of the semantic communication information through the global model.

[0347] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0348] If the semantic processing is successful, the semantic information is obtained.

[0349] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0350] According to the semantic information, it is determined whether the current semantic communication process is completed.

[0351] If the current semantic communication process is completed, semantic communication completion information is sent to the second terminal.

[0352] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0353] An operation corresponding to the semantic information is performed.

[0354] In one embodiment, the computer program, when executed by the processor, specifically implements the following steps:

[0355] It is determined whether the semantic information meets a communication intention or a communication target corresponding to the semantic communication information.

[0356] If the semantic information meets the communication intention or the communication target, it is determined that the current semantic communication process is completed.

[0357] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0358] If the current semantic communication process is not completed, feedback information is sent to the second terminal, wherein the feedback information is used to instruct the second terminal to re-execute the semantic communication process.

[0359] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0360] receiving the semantic communication information re-sent by the second terminal, and performing semantic processing on the semantic communication information re-sent by the second terminal to obtain semantic information.

[0361] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0362] receiving the updated model parameters sent by the server, the updated model parameters being obtained after the server performs semantic processing on the approximate feature to obtain semantic information and then receives semantic communication completion information sent by the second terminal, and performing model parameter updating on the global model according to the approximate feature to obtain the updated model parameters;

[0363] updating the local semantic knowledge base model based on the updated model parameters.

[0364] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0365] receiving the semantic processing failure feedback sent by the server;

[0366] sending the semantic processing failure feedback to the second terminal.

[0367] The embodiment of the application further provides a computer program product containing instructions, the computer program being executed by the processor to implement the following steps:

[0368] receiving the semantic communication information sent by the first terminal, and performing semantic processing on the semantic communication information;

[0369] sending the semantic information obtained by the semantic processing to the first terminal.

[0370] In one embodiment, the computer program, when executed by the processor, specifically implements the following steps:

[0371] performing semantic processing on the semantic communication information according to the global model;

[0372] if the semantic information is not obtained, performing information disassembly on the semantic communication information to obtain an approximate feature, and performing semantic processing on the approximate feature through the global model.

[0373] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0374] receiving the semantic communication completion information sent by the second terminal, wherein the semantic communication completion information is determined by the first terminal according to semantic information, and the semantic information is obtained by the server by performing semantic processing on an approximate feature through a global model;

[0375] performing model parameter updating on the global model according to the approximate feature to obtain updated model parameters;

[0376] send the updated model parameters to the first terminal.

[0377] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0378] If semantic information is not obtained in the process of performing semantic processing on the approximate feature by the global model, a semantic processing failure feedback is sent to the first terminal.

[0379] In one embodiment, the computer program, when executed by the processor, further implements the following steps:

[0380] If a plurality of failure notifications sent by the second terminal are received, the approximate feature obtained by disassembling is deleted.

[0381] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0382] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0383] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the scope of protection of the present application. Therefore, the scope of protection of the patent of the present application should be subject to the appended claims.

Claims

A semantic communication method, comprising: performing semantic processing on semantic communication information in a semantic communication process; if the semantic processing fails, sending the semantic communication information to a server to process the semantic communication information and obtain semantic information corresponding to the semantic communication information, and returning the semantic information. According to the method of claim 1, wherein the semantic processing comprises semantic extraction and / or semantic restoration. According to the method of claim 1, wherein the semantic processing on the semantic communication information in the semantic communication process comprises: performing semantic processing on the semantic communication information by using a local semantic knowledge base model. According to the method of claim 3, wherein the local semantic knowledge base model is a semantic knowledge base base model constructed by a first terminal through federated learning. According to the method of claim 3, the method further comprises: receiving the semantic communication information sent by a second terminal. According to the method of claim 5, the method further comprises: receiving the semantic information issued by the server; wherein the semantic information is obtained by the server performing semantic processing on the semantic communication information by using a global model, or the semantic information is obtained by the server performing semantic processing on approximate features of the semantic communication information by using the global model. According to the method of claim 5, the method further comprises: if the semantic processing succeeds, obtaining the semantic information. According to the method of claim 6 or 7, the method further comprises: determining whether the current semantic communication process is completed according to the semantic information; if it is determined that the current semantic communication process is completed, sending semantic communication completion information to the second terminal. According to the method of claim 8, wherein after the semantic communication completion information is sent to the second terminal, the method further comprises: performing an operation corresponding to the semantic information. According to the method of claim 8, wherein the determining whether the current semantic communication process is completed according to the semantic information comprises: determining whether the semantic information meets a communication intent or a communication target corresponding to the semantic communication information; if it is determined that the semantic information meets the communication intent or the communication target, it is determined that the current semantic communication process is completed. According to the method of claim 8, the method further comprises: if it is determined that the current semantic communication process is not completed, sending feedback information to the second terminal, the feedback information being used to instruct the second terminal to re-execute the semantic communication process. According to the method of claim 11, the method further comprises: receiving the semantic communication information re-sent by the second terminal, and performing semantic processing on the semantic communication information re-sent by the second terminal to obtain the semantic information. According to the method of claim 6, the method further comprises: receive the updated model parameters sent by the server, the updated model parameters being obtained by the server after semantic processing of the approximate features and before receiving semantic communication completion information sent by the second terminal, and the updated model parameters being obtained by updating model parameters of the global model according to the approximate features; update the local semantic knowledge base model based on the updated model parameters. The method of claim 1, further comprising: receiving semantic processing failure feedback sent by the server; sending the semantic processing failure feedback to a second terminal. A semantic communication method, comprising: receiving semantic communication information sent by a first terminal and performing semantic processing on the semantic communication information; sending semantic information obtained by the semantic processing to the first terminal. The method of claim 15, wherein the semantic processing on the semantic communication information comprises: performing semantic processing on the semantic communication information according to a global model; if semantic information is not obtained, performing information disassembly on the semantic communication information to obtain approximate features, and performing semantic processing on the approximate features by the global model. The method of claim 16, further comprising: receiving semantic communication completion information sent by a second terminal, wherein the semantic communication completion information is determined by the first terminal according to semantic information obtained by the server by performing semantic processing on the approximate features by the global model; updating model parameters of the global model according to the approximate features to obtain updated model parameters; sending the updated model parameters to the first terminal. The method of claim 16, further comprising: if semantic information is not obtained in the process of performing semantic processing on the approximate features by the global model, sending semantic processing failure feedback to the first terminal. The method of claim 18, further comprising: if receiving multiple failure notifications sent by the second terminal, deleting the approximate features obtained by the disassembly. A semantic communication apparatus, comprising: a semantic processing module configured to perform semantic processing on semantic communication information in a semantic communication process; a sending module configured to, if the semantic processing fails, send the semantic communication information to a server to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and return the semantic information. A semantic communication apparatus, comprising: a receiving module configured to receive semantic communication information sent by a first terminal and perform semantic processing on the semantic communication information; a sending module configured to send semantic information obtained by the semantic processing to the first terminal. A terminal, comprising a transmitter and a processor; the processor is configured to perform semantic processing on semantic communication information in a semantic communication process; the transmitter is configured to, if the semantic processing fails, send the semantic communication information to a server to process the semantic communication information by the server and obtain semantic information corresponding to the semantic communication information, and return the semantic information. A server comprising a transmitter, a processor and a receiver: the receiver configured to receive semantic communication information transmitted by a first terminal; the processor configured to perform semantic processing on the semantic communication information; the transmitter configured to transmit semantic information obtained by the semantic processing to the first terminal. A semantic communication system comprising a first terminal, a second terminal and a server; the second terminal configured to transmit semantic communication information to the first terminal; the first terminal configured to perform the steps of the method according to any one of claims 1-14; the server configured to perform the steps of the method according to any one of claims 15-19. The semantic communication system according to claim 24, wherein the second terminal is further configured to perform a knowledge base search according to the communication information to be transmitted to obtain the semantic communication information; the second terminal is further configured to, if the knowledge base search fails, transmit the communication information to be transmitted to the server, so that the server performs a knowledge base search according to the communication information to be transmitted to obtain the semantic communication information. A computer readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the steps of the method according to any one of claims 1-19. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of the method according to any one of claims 1-19.

Images