Object recognition method and device, electronic equipment, storage medium and program product

Abstract

The embodiment of the invention provides an object recognition method and device, electronic equipment, a storage medium and a program product. The method comprises the steps of obtaining a second signal list of at least one first historical object in response to an object recognition request based on a first signal list; determining the similarity between the first signal list and the second signal list according to the correlation between a first signal in the first signal list and a different second signal in the second signal list; and outputting an object recognition result corresponding to the first signal list based on the similarity. According to the embodiment of the invention, by utilizing the technical scheme, the condition of misrecognition of the object can be reduced, and the accuracy of the generated object recognition result is improved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of object recognition, and in particular, to an object recognition method and device, electronic device, storage medium, and program product. BACKGROUND

[0002] In some scenarios, there is a demand for object recognition, such as a demand for identifying a point of interest visited by a user based on a request of the user. However, in the related art, the accuracy of the generated object recognition result is low. SUMMARY

[0003] Embodiments of the present disclosure provide an object recognition method, device, electronic device, storage medium, and program product to improve the accuracy of the generated object recognition result.

[0004] In a first aspect, embodiments of the present disclosure provide an object recognition method, comprising:

[0005] In response to an object recognition request based on a first signal list, obtaining a second signal list of at least one first historical object;

[0006] According to the correlation between a first signal in the first signal list and different second signals in the second signal list, determining the similarity between the first signal list and the second signal list;

[0007] Based on the similarity, outputting an object recognition result corresponding to the first signal list.

[0008] In a second aspect, embodiments of the present disclosure also provide an object recognition device, comprising:

[0009] A list obtaining module configured to, in response to an object recognition request based on a first signal list, obtain a second signal list of at least one first historical object;

[0010] A similarity determining module configured to, according to the correlation between a first signal in the first signal list and different second signals in the second signal list, determine the similarity between the first signal list and the second signal list;

[0011] A result outputting module configured to, based on the similarity, output an object recognition result corresponding to the first signal list.

[0012] In a third aspect, embodiments of the present disclosure also provide an electronic device, comprising:

[0013] One or more processors;

[0014] A memory configured to store one or more programs,

[0015] When the one or more programs are executed by the one or more processors, the one or more processors implement the object identification method according to the embodiments of the present disclosure.

[0016] In a fourth aspect, the embodiments of the present disclosure further provide a computer-readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the object identification method according to the embodiments of the present disclosure.

[0017] In a fifth aspect, the embodiments of the present disclosure further provide a computer program product, which, when executed by a computer, causes the computer to implement the object identification method according to the embodiments of the present disclosure.

[0018] The object identification method, device, electronic device, storage medium and program product provided by the embodiments of the present disclosure can reduce the situation of object misrecognition and improve the accuracy of the generated object identification result by comprehensively considering the correlation between the first signals in the first signal list and different second signals in the second signal list when identifying the object based on the first signal list. BRIEF DESCRIPTION OF DRAWINGS

[0019] The above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent by describing in detail the embodiments thereof with reference to the attached drawings. The same or similar components are denoted by the same or similar reference numerals throughout the drawings. It is to be understood that the drawings are schematic and elements are not necessarily drawn to scale.

[0020] Figure 1 A flowchart of an object identification method according to an embodiment of the present disclosure;

[0021] Figure 2 A flowchart of an object identification method according to an embodiment of the present disclosure;

[0022] Figure 3 A result diagram of a model according to an embodiment of the present disclosure;

[0023] Figure 4 A structural block diagram of an object identification device according to an embodiment of the present disclosure;

[0024] Figure 5 A structural diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be embodied in various forms and should not be interpreted as being limited to the embodiments set forth herein; rather, these embodiments are provided so as to more completely and thoroughly understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for illustrative purposes and are not intended to limit the scope of protection of the present disclosure.

[0026] It should be understood that each of the steps recited in the method embodiments of the present disclosure can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0027] The term "comprising" and variations thereof as used herein are open-ended, that is "including but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Related terms are defined in the following description.

[0028] It should be noted that the terms "first", "second", and the like in the present disclosure are only used to distinguish different devices, modules or units, and are not intended to limit the order or interdependence of the functions performed by these devices, modules or units.

[0029] It should be noted that the terms "one", "multiple" in the present disclosure are illustrative and not limiting, and those skilled in the art should understand that unless otherwise explicitly stated in the context, it should be understood as "one or more".

[0030] The names of the messages or information exchanged between the devices in the embodiments of the present disclosure are only for illustrative purposes, and are not intended to limit the scope of the messages or information.

[0031] It can be understood that before using the technical solutions disclosed in the embodiments of the present disclosure, the type, scope of use, use scenario, etc. of the personal information involved in the present disclosure should be informed to the user and the authorization of the user should be obtained in accordance with relevant laws and regulations.

[0032] For example, when responding to the active request of the user, a prompt message is sent to the user to explicitly prompt the user that the operation requested to be performed will require the acquisition and use of the personal information of the user. Thus, the user can voluntarily choose whether to provide personal information to the software or hardware, such as electronic devices, application programs, servers or storage media, etc. that perform the operation of the technical solutions of the present disclosure according to the prompt message.

[0033] As an optional but non-limiting implementation, in response to receiving the active request of the user, the manner of sending the prompt information to the user may, for example, be a pop-up window manner, in which the prompt information can be presented in a textual manner. In addition, the pop-up window can also carry a selection control for the user to select "agree" or "disagree" to provide personal information to the electronic device.

[0034] It can be understood that the above notification and user authorization obtaining process is only illustrative and does not limit the implementation of the present disclosure. Other manners that meet relevant laws and regulations can also be applied to the implementation of the present disclosure.

[0035] It can be understood that the data involved in the technical solution (including but not limited to the data itself, the acquisition or use of the data) should comply with the requirements of the relevant laws and regulations and the relevant provisions.

[0036] Figure 1 A flowchart of an object identification method provided by the embodiment of the present disclosure is shown. The method can be executed by an object identification device, which can be implemented by software and / or hardware and can be configured in an electronic device, typically, in a computer, a mobile phone or a tablet computer. The object identification method provided by the embodiment of the present disclosure is suitable for a scenario of identifying an object based on a signal list, such as a scenario of identifying a visiting interest point of a user according to a scanned wireless network communication WIFI signal based on an active request of the user. Figure 1 As shown, the object identification method provided by the embodiment can include:

[0037] S101, in response to an object identification request based on a first signal list, obtaining a second signal list of at least one first historical object.

[0038] The object identification request can be understood as a request for identifying an object corresponding to a signal list, such as a request for identifying an object matched with a signal list. The object can be understood as an object to be identified, such as a visiting interest point, a device and / or an article to be identified. The first signal list can be a signal list corresponding to the current object identification request to be processed, in other words, the signal list for which the current object identification request requests to identify an object. The first signal list includes a plurality of first signals, which can be understood as signals for identifying an object, such as signals that can be received and / or detected by a client or an electronic device sending an object identification request. Taking the scenario of identifying a visiting interest point as an example, the signal can be a signal sent by a corresponding device configured in each interest point and specially used for identifying the object; or can be a wireless network communication signal in each interest point, such as a WIFI signal that can be scanned by each interest point.

[0039] The first historical object can be understood as a historical object to be matched with the first signal list, such as a candidate object that the first signal list can correspond to. This historical object can be understood as an object whose corresponding signal list has been determined by history, which can include an object whose corresponding signal list has been obtained by the signal collection of the collection personnel. In the first example, this historical object can also include an object whose identification is determined by history in the case of active authorization of the relevant personnel of the request identification side and / or the object side of the corresponding object. The second signal list can be understood as a list of second signals scanned at the first historical object, such as a list of signals that can be detected or collected at the first historical object. The second signal object can include a plurality of signals that can be detected or collected at the first historical object, that is, a plurality of second signals. Taking the first historical object as a point of interest and the above-mentioned signal as a WIFI signal as an example, the second signal can be a WIFI signal that can be scanned in the point of interest, and the second signal list can be a list of WIFI signals that can be scanned in the point of interest.

[0040] The first signal list can be understood as a list of first signals. The first signal list can include one or more first signals, which can be determined according to the signal scanning condition of the client or electronic device sending the object identification request. The second signal list of the first historical object can include one or more second signals, which can be determined according to the signal scanning condition at the first historical object. The following is described by taking the first signal list including a plurality of first signals and the second signal list including a plurality of second signals as an example. The first signal list can record signal information of at least part of the first signals; the second signal list of a certain first historical object can record signal information of at least part of the second signals that can be scanned at the first historical object.

[0041] The specific content included in the signal information is not limited, and exemplarily, the signal information can include at least one of the following: identification information of the signal, object association information, signal strength distribution information, first sorting distribution information, and second sorting distribution information.

[0042] The identification information can be understood as unique identity information of the signal, such as a media access control (Media Access Control, MAC) address corresponding to the signal. The MAC address can be the MAC address of the device sending the signal, such as the MAC address of the routing device sending the WIFI signal.

[0043] The object association information can be understood as information indicating whether there is an associated object (such as a point of interest) in the signal and what the associated object is. The determination manner of the object association information of the signal is not limited. For example, in the visiting point of interest identification scene based on the WIFI signal list, the WIFI signal associated with the point of interest can be determined based on the report of the point of interest party; or the object association information of the WIFI signal can be determined based on the similarity between the name of the WIFI signal (such as the name of the routing device sending the WIFI signal) and the name of the point of interest, and the like.

[0044] The signal strength distribution information can be understood as information describing the distribution of the signal strength, for example, information describing the distribution of the signal in different signal strength intervals, such as the distribution times in different signal strength intervals. For example, for a signal with only one acquisition (such as the first signal), the signal strength can be 100% distributed in a certain signal strength interval, for example, the distribution times in the signal strength interval is 1, and the distribution times in other signal strength intervals except the signal strength interval is 0; for a signal with multiple acquisitions (such as the second signal), the signal strength can be distributed in one or more signal strength intervals, for example, the distribution times in one or more signal strength intervals can be greater than or equal to 1.

[0045] The first sorting distribution information of a signal can be understood as the arrangement times of the signal in different first position intervals after sorting at least one signal collected at the corresponding object according to the signal strength. For example, in the visiting point of interest identification scene based on the WIFI signal list, the first sorting distribution information can be the historical visiting index position of the WIFI signal, for example, the position distribution information of the WIFI signal arranged according to the signal strength in the WIFI signal scanned by each person visiting the corresponding point of interest. For example, for a signal with only one acquisition (such as the first signal), the signal can be 100% distributed in a certain first position interval, for example, the distribution times in the first position interval is 1, and the sorting position in other first position intervals except the first position interval is 0; for a signal with multiple acquisitions (such as the second signal), the signal can be distributed in one or more first position intervals, for example, the distribution times in one or more first position intervals can be greater than or equal to 1. It should be noted that the signals collected at each object are obtained on the premise of active authorization of the relevant person.

[0046] The second sorting distribution information of a signal can be understood as the arrangement times of the signal in different second position intervals after sorting different signals scanned in each signal scanning according to signal strength. Taking the visited point of interest identification scene based on the WIFI signal list as an example, the second sorting distribution information can be the full collection signal strength distribution of the WIFI signal, for example, the position distribution information of the WIFI signal arranged according to the signal strength in the historical object identification request in which the WIFI signal is scanned. The division manner of the first position interval and the division manner of the second position interval can be the same or different. It should be noted that the signal strength sorting in the signal list carried in each historical object identification request is obtained under the premise of active authorization by relevant personnel.

[0047] For example, in the case of receiving an object identification request for a certain signal list (such as a first signal list), at least one second signal list can be obtained, and different second signal lists can be associated with different historical objects, such as different first historical objects. For example, in the case of receiving an object identification request for a first signal list, one or more signal lists (i.e., second signal lists) of the first historical objects can be obtained from the object library in response to the object identification request.

[0048] In some embodiments, the first signal list can not be considered, and the second signal list of one or more historical objects can be directly obtained, such as the signal list of all historical objects in the object library, which is directly obtained as the second signal list; or, according to a set condition, the signal list of part of the historical objects in the object library is obtained, such as the signal list of the historical objects located in a geographic location area (such as a province, a city, or a region, etc.) where the electronic device sending the object identification request is located, which is obtained from the object library as the second signal list, etc.

[0049] In some embodiments, the first signal list can be considered, such as obtaining the signal list of one or more historical objects from the object library according to the first signal list as the second signal list, so as to further improve the accuracy and practicability of the obtained second signal list. In this case, optionally, the obtaining of the second signal list of the at least one first historical object comprises: obtaining the second signal list of the at least one first historical object according to the first signal list.

[0050] In the above embodiments, the manner of obtaining the second signal list according to the first signal list can be set as required, for example, one or more signal lists of historical objects can be obtained from the object library according to the number of repeated signals between each signal list in the object library and the first signal list, such as a signal list having repeated signals with the first signal list, or a signal list having a number of repeated signals satisfying a preset number condition with the first signal list, as the second signal list, and the like.

[0051] Optionally, the obtaining of the second signal list of the at least one first historical object according to the first signal list comprises: obtaining at least one third signal list having repeated signals with the first signal list from an object library, wherein the object library stores a plurality of signal lists of historical objects; obtaining at least part of the third signal list as the second signal list from the at least one third signal list, and taking the historical object corresponding to the second signal list as the first historical object, wherein the number of repeated signals between the at least part of the third signal list and the first signal list satisfies a preset number condition.

[0052] The third signal list can be understood as a signal list of an object stored in the object library and having repeated signals with the first signal list. The object library can be a database for storing signal lists of historical objects. For example, the object library can store a plurality of signal lists of objects, and the number of signal lists of a single object can be one or more, which is not limited in the embodiment. The repeated signals are repeated signals, such as identical signals. The preset number condition can be set as required, for example, the preset number condition can include that the number of repeated signals is greater than or equal to a preset number threshold, and / or is located in the top M positions in the sorting according to the number of repeated objects from more to less, and the like. The preset number threshold and M can be set as required, for example, the preset number threshold can be set as 2, 3, 5, or the like, and M can be set as 20, 30, 40, or the like.

[0053] For example, a signal list having repeated signals with the first signal list can be obtained from the object library as the third signal list. For example, a signal list including a first signal in the first signal list can be obtained from the object library as a signal list corresponding to the first signal, and after obtaining the signal list corresponding to each first signal or at least part of the first signal in the first signal list, a de-duplication processing can be performed on the signal list corresponding to each first signal, such as removing the repeated signal list, and the signal list after the de-duplication processing is taken as the third signal list.

[0054] After obtaining the third signal lists, a third signal list in which the number of repeated signals between the first signal list and the second signal list satisfies a preset number condition can be further selected from the obtained third signal lists as the second signal list. For example, if the number of the obtained third signal lists is greater than M, the third signal lists can be sorted in order of the number of repeated signals between the first signal list and the second signal list from more to less, and the third signal lists located in the first M positions in the sorting are obtained as the second signal list; if the number of the obtained second signal lists is less than or equal to M, the obtained third signal lists can all be taken as the second signal list. Further, the historical object corresponding to each determined second signal list can be taken as the first historical object.

[0055] In the embodiment, the arrangement manner of each first signal in the first signal list is not limited to the arrangement manner of each second signal in the second signal list. In some examples, each first signal in the first signal list and each second signal in the second signal list can be sorted according to signal strength information, such as arranged in order of signal strength evaluation value from large to small or from small to large, etc., and the following is described by taking the arrangement in order of signal strength evaluation value from large to small as an example. The signal strength evaluation value may, for example, be the average, median or mode of the signal strength, etc., which can be set according to needs.

[0056] S102, determining the similarity between the first signal list and the second signal list according to the correlation between the first signal in the first signal list and different second signals in the second signal list.

[0057] In the embodiment, after obtaining at least one second signal list, the correlation between one or more first signals in the first signal list and different second signals in the second signal list can be determined for each second signal list or part of the second signal list, and the similarity between the first signal list and the second signal list is determined according to the correlation.

[0058] In some embodiments, the similarity between the first signal list and each second signal list can be determined based on a pre-set similarity determination manner. For example, a feature vector of at least part of the first signals in the first signal list and / or a feature vector of at least part of the second signals in the second signal list can be generated based on a correlation score between the at least part of the first signals in the first signal list and the different second signals in the second signal list. For example, the original feature vector of the at least part of the first signals in the first signal list and / or the original feature vector of the at least part of the second signals in the second signal list can be updated based on the correlation score to obtain an updated feature vector of the corresponding signals. Then, the similarity between the first signal list and the second signal list can be calculated based on the feature vector of the at least part of the first signals in the first signal list and the feature vector of the at least part of the second signals in the second signal list.

[0059] In some embodiments, a pre-trained model can be used to determine the similarity between the first signal list and each second signal list. For example, the first signal list and each second signal list can be input into the model, and the similarity between the first signal list and each second signal list output by the model can be obtained. In this way, the resources consumed for feature vector extraction and maintenance can be reduced, different application scenarios can be quickly migrated, and the generalization of the model can be improved. In this case, the similarity between the first signal list and the second signal list can be determined based on the correlation between the first signals in the first signal list and the different second signals in the second signal list, which can include inputting the first signal list and the second signal list into a pre-trained model, and determining the similarity between the first signal list and the second signal list based on the correlation between the first signals in the first signal list and the different second signals in the second signal list by the model. The model can be a model that supports the calculation of the similarity between signal lists, and the type of the model is not limited. For example, the model can be a large language model, or a model other than a large language model, such as a Bidirectional Encoder Representations from Transformers (BERT) model, etc.

[0060] S103、based on the similarity, output an object recognition result corresponding to the first signal list.

[0061] Specifically, after obtaining the similarity between the first signal list and each second signal list, an object recognition result of the first signal list can be generated based on the similarity. For example, a first historical object corresponding to the second signal list with the highest similarity to the first signal list can be determined as an object (e.g., a visited point of interest) corresponding to the first signal list, and an object recognition result corresponding to the first signal list can be generated based on this.

[0062] In some embodiments, the first historical object corresponding to the second signal list with the highest similarity to the first signal list can be directly obtained as the object corresponding to the first signal list without considering the specific size of the similarity between the second signal list with the highest similarity to the first signal list and the first signal list, so as to reduce the operations required for generating the object recognition result and further improve the generation speed of the object recognition result.

[0063] In some embodiments, the specific size of the similarity between the second signal list with the highest similarity to the first signal list and the first signal list can be considered, such as the relative size between the similarity and a preset similarity threshold, to generate different object recognition results, so as to further improve the accuracy of the generated object recognition result.

[0064] Optionally, the output of the object recognition result corresponding to the first signal list based on the similarity includes at least one of the following: in response to the presence of a second historical object in the at least one first historical object, outputting a first object recognition result corresponding to the first signal list based on the second historical object, wherein the similarity between the second signal list of the second historical object and the first signal list satisfies a preset similarity condition; and in response to the absence of a second historical object in the at least one first historical object, outputting a second object recognition result corresponding to the first signal list.

[0065] The second historical object can be an object corresponding to a second signal list with a similarity to the first signal list satisfying a preset similarity condition, i.e., an object determined based on the first signal list. The preset similarity condition can be set as needed, for example, the preset similarity condition can include the highest similarity and the similarity being greater than or equal to a preset similarity threshold, etc. The specific value of the preset similarity threshold is not limited, for example, the preset similarity threshold can be 0.5, 0.6, or 0.7, etc. The first object recognition result can be an object recognition result generated when the second historical object exists. For example, the first object recognition result can include object information (such as identification information or item information, etc.) of the second historical object and / or recommendation information associated with the second historical object, etc. The second object recognition result can be an object recognition result generated when the second historical object does not exist. The second object recognition result can include prompt information, which can be used to prompt that the object corresponding to the first signal list is not recognized.

[0066] Specifically, it can be determined whether there is a second historical object that satisfies a preset similarity condition based on the similarity between the second signal list of each first historical object and the first signal list. For example, the second signal list with the highest similarity to the first signal list can be determined, and it is further determined whether the similarity between the second signal list and the first signal list is greater than or equal to a preset similarity. If yes, the first historical object corresponding to the second signal list is taken as the second historical object, and it is determined that there is a second historical object. If no, it is determined that there is no second historical object.

[0067] If there is a second historical object, the second historical object can be determined as an object identified based on the first signal list, a first signal identification result corresponding to the first signal list is generated based on the second historical object, and the first signal identification result is output, such as displaying information associated with the first historical object to the user.

[0068] If there is no second historical object, a second object identification result corresponding to the first signal list can be generated and output, such as prompting the user that no object corresponding to the first signal list is identified.

[0069] The object identification method provided in the embodiment responds to an object identification request based on a first signal list, acquires a second signal list of at least one first historical object, determines the similarity between the first signal list and each second signal list based on the correlation between the first signal in the first signal list and different second signals in the second signal list, and outputs an object identification result corresponding to the first signal list based on the similarity. The above technical solution is used in the embodiment. When object identification is performed based on the first signal list, the correlation between the first signal in the first signal list and different second signals in the second signal list is comprehensively considered, which can reduce the situation of object misidentification and improve the accuracy of the generated object identification result.

[0070] Figure 2Another flowchart of an object recognition method is provided in the embodiments of the present disclosure. The solution in the embodiments can be combined with one or more optional solutions in the above embodiments. Optionally, the similarity between the first signal list and the second signal list is determined according to the correlation between the first signals in the first signal list and different second signals in the second signal list, including: generating a first feature vector sequence between the first signal list and the second signal list, the first feature vector sequence including feature vectors of a plurality of signals, the plurality of signals including at least part of the first signals in the first signal list and at least part of the second signals in the second signal list; updating at least part of the feature vectors in the first feature vector sequence according to the correlation between different feature vectors in the first feature vector sequence to obtain a second feature vector sequence; and determining the similarity between the first signal list and the second signal list based on the second feature vector sequence.

[0071] Correspondingly, as shown in Figure 2 The object recognition method provided in the embodiments can include:

[0072] S201, in response to an object recognition request based on a first signal list, obtaining a second signal list of at least one first historical object.

[0073] S202, generating a first feature vector sequence between the first signal list and the second signal list, the first feature vector sequence including feature vectors of a plurality of signals, the plurality of signals including at least part of the first signals in the first signal list and at least part of the second signals in the second signal list.

[0074] In the embodiments, after obtaining the first signal list and at least one second signal list, a first feature vector sequence between the first signal list and each second signal list can be generated, such as generating a first feature vector sequence between the first signal and the second signal list based on the feature vectors of at least part of the first signals in the first signal list and the feature vectors of at least part of the second signals in the second signal list.

[0075] The first feature vector sequence between the first signal list and the second signal list can be understood as a feature vector sequence initially generated based on the feature vectors of the first signals in the first signal list and the feature vectors of the second signals in the second signal list. For example, the correlation between the first signals in the first signal list and different second signals in the second signal list has not been considered when generating the first feature vector sequence.

[0076] The generation manner of the first eigenvector sequence is not limited. Optionally, the signal information of at least one first signal in the first signal list, the signal information of at least one second signal in the second signal list, and the generation of the first eigenvector sequence between the first signal list and the second signal list include: generating the eigenvector of the at least one first signal according to the signal information of the at least one first signal in the first signal list; and generating the eigenvector of the at least one second signal according to the signal information of the at least one second signal in the second signal list; and constructing the first eigenvector sequence between the first signal list and the second signal list based on the eigenvector of the at least one first signal and the eigenvector of the at least one second signal.

[0077] For example, for each of the at least one first signal and the at least one second signal, the signal information of the signal in at least one dimension is obtained from the signal list corresponding to the signal, and the eigenvector of the signal is generated based on the signal information of the signal in the at least one dimension, such as respectively encoding the signal information in the at least one dimension to obtain the encoding vector of the signal in the at least one dimension; and the encoding vectors of the signal in the at least one dimension are added, and the added vector is taken as the eigenvector of the signal.

[0078] After obtaining the eigenvectors of the at least one first signal, the eigenvectors of the at least one first signal can be sorted according to the arrangement order of the at least one first signal in the first signal list (such as the order from high to low of the signal strength evaluation value, etc.), to obtain the eigenvector sequence corresponding to the at least one first signal; and after obtaining the eigenvectors of the at least one second signal, the eigenvectors of the at least one second signal can be sorted according to the arrangement order of the at least one second signal in the second signal list (such as the order from high to low of the signal strength evaluation value, etc.), to obtain the eigenvector sequence corresponding to the at least one second signal. Then, the eigenvector sequence corresponding to the at least one second signal can be spliced before or after the eigenvector sequence corresponding to the at least one first signal, and the spliced eigenvector sequence is taken as the first eigenvector sequence between the first signal list and the second signal list.

[0079] When encoding the signal information in different dimensions, the encoding processing manner can be the same or different, which can be set according to needs.

[0080] For example, the signal information includes at least five dimensions of signal identification information, object association information, signal strength distribution information, first order distribution information and second order distribution information. The signal identification information can be encoded by using a relative index coding method. For example, the MAC address of the signal is first indexed and coded based on the preset address segment length, and the MAC address of the signal is mapped to a vector space to obtain the encoding vector of the MAC address. The object association information is encoded, for example, by using 0 and 1 to indicate whether the signal is a candidate object (such as each first historical object). The signal strength distribution is discretely binned (such as determining the distribution times in each signal strength interval), and is mapped to a vector space by using a linear mapping method. The first order distribution information of the signal is discretely binned (such as determining the distribution times in each first bit interval), and is mapped to a vector space by using a linear mapping method. And / or, the second order distribution information of the signal is discretely binned (such as determining the distribution times in each second bit interval), and is mapped to a vector space by using a linear mapping method, and the like.

[0081] S203, updating at least part of the feature vectors in the first feature vector sequence according to the correlation between different feature vectors in the first feature vector sequence, to obtain a second feature vector sequence.

[0082] The second feature vector sequence can be understood as a feature vector sequence obtained by updating the first feature vector sequence based on the correlation between the feature vectors, such as the updated first feature vector sequence.

[0083] In this step, at least part of the feature vectors in the first feature vector sequence can be updated according to the correlation scores between different feature vectors in the first feature vector sequence. For example, the feature vectors of at least part of the first signals and / or the feature vectors of at least part of the second signals in the first feature vector sequence are updated based on the correlation between the feature vectors of each first signal and the feature vectors of each second signal in the first feature vector sequence. Alternatively, part or all of the feature vectors in the first feature vector sequence are updated based on the correlation between each pair of feature vectors in the first feature vector sequence, and the like, thereby obtaining the second feature vector sequence.

[0084] In this embodiment, the updating manner of the feature vectors in the first feature vector sequence is not limited. In some examples, a preset attention formula can be used to perform attention calculation on the first feature vector sequence to obtain a correlation score between different feature vectors in the first feature vector sequence as a calculation result of the attention calculation; then, at least part of the feature vectors in the first feature vector sequence can be updated based on the calculation result to obtain the second feature vector sequence. In this case, the updating of the at least part of the feature vectors in the first feature vector sequence according to the correlation between different feature vectors in the first feature vector sequence can include: using a preset attention calculation formula to perform attention calculation on the feature vectors in the first feature vector sequence; and updating at least part of the feature vectors in the first feature vector sequence based on a calculation result of the attention calculation.

[0085] The preset attention calculation formula can include or not include a bias term matrix. For example, the preset attention calculation formula can include a bias term matrix to further improve the accuracy of the attention calculation. The bias term matrix can be set as needed. Optionally, the bias term matrix in the preset attention calculation formula is associated with difference information between two of the plurality of signals, and the difference information includes first difference information of signal intensity distribution and / or second difference information of signal intensity order distribution.

[0086] For example, the preset attention calculation formula can be:

[0087]

[0088] In the formula, Q represents a Query matrix, K represents a Key matrix, and V represents a Value matrix. is a scaling factor used to prevent the gradient of the softmax function from disappearing due to an excessively large inner product. pos is a bias term matrix, which can be calculated based on difference information of signal intensity distribution information (i.e., first difference information) between two of different signals and / or difference information of signal intensity order distribution information (e.g., first order distribution information or second order distribution information) between two of different signals (i.e., second difference information). For example, when the bias term matrix is calculated based on the difference information of the signal intensity distribution information between two of different signals and the difference information of the first order distribution information between two of different signals, the calculation formula can be:

[0089] W pos = W rssi α rssi + W index α index ,

[0090] In the above formula, W rssi is a bias item encoding matrix corresponding to the signal intensity distribution information; W index is a bias item encoding matrix corresponding to the first ranking distribution information, W rssi is different from W index may be pre-set or obtained through training, and is used to encode the positional relationship between different signals, such as the relationship of the signal intensity distribution information or the relationship of the first ranking distribution information between different signals. rssi is a matrix of difference information of the signal intensity distribution information between different signals, such as a signal intensity bias matrix, α index is a matrix of difference information of the first ranking distribution information between different signals, that is, an index bit offset matrix, in some examples, α rssi is different from α index may be characterized by the Wasserstein distance. In this embodiment, the weights can be normalized by the softmax function to ensure that the sum of all weights is 1.

[0091] S204, determining the similarity between the first signal list and the second signal list based on the second feature vector sequence.

[0092] In this embodiment, after the second feature vector sequence is updated, the similarity between the first signal list and the second signal list can be determined based on the second feature vector sequence.

[0093] In some embodiments, the first feature vector sequence can not be considered, and the similarity between the first signal list and the second signal list can be directly determined based on the updated second feature vector sequence.

[0094] For example, the feature vectors of each first signal and the feature vectors of each second signal in the second feature vector sequence can be determined, and for each first signal, the similarity between the first signal and each second signal can be calculated based on the feature vector of the first signal in the second feature vector sequence and the feature vectors of each second signal in the second feature vector sequence. After calculating the similarity between each first signal and each second signal based on the second feature vector sequence, the average or weighted average of the similarities can be calculated as the similarity between the first signal list and the second signal list.

[0095] For example, the feature vectors of each first signal in the second feature vector sequence and the feature vectors of each second signal can be determined, and for each first signal, the similarity between the first signal and a second signal in the second signal list having the same rank as the first signal in the first signal list can be calculated based on the feature vector of the first signal in the second feature vector sequence and the feature vector of the second signal in the second signal list. After the similarity between each first signal and the second signal having the same rank as the first signal is calculated based on the second feature vector sequence, the average or weighted average of the similarities can be calculated as the similarity between the first signal list and the second signal list.

[0096] For another example, the feature vectors of the first signal list can be generated based on the feature vectors of each first signal in the second feature vector sequence, such as concatenating the feature vectors of each first signal in the order of their arrangement in the second feature vector sequence to obtain the feature vectors of the first signal list. In addition, the feature vectors of the second signal list can be generated based on the feature vectors of each second signal in the second feature vector sequence, such as concatenating the feature vectors of each second signal in the order of their arrangement in the second feature vector sequence to obtain the feature vectors of the second signal list. Then, the similarity between the feature vectors of the first signal list and the feature vectors of the second signal list can be calculated as the similarity between the first signal list and the second signal list.

[0097] In some embodiments, the first feature vector sequence can be considered, for example, the similarity between the first signal list and the second signal list can be determined based on the first feature vector sequence and the second feature vector sequence obtained by updating at least part of the feature vectors in the first feature vector sequence, so as to further improve the accuracy of the determined similarity and further improve the accuracy of the generated object recognition result.

[0098] Optionally, the determining the similarity between the first signal list and the second signal list based on the second feature vector sequence comprises: determining the similarity between the first signal list and the second signal list based on the first feature vector sequence and the second feature vector sequence.

[0099] Exemplarily, the first feature vector sequence and the second feature vector sequence can be combined to obtain a third feature vector sequence, such as adding each feature vector in the first feature vector sequence to a feature vector at a same arrangement position in the second feature vector sequence, and taking a feature vector sequence obtained by the addition of the feature vectors as the third feature vector sequence. After obtaining the third feature vector sequence, the similarity between the first signal list and the second signal list can be determined based on the third feature vector sequence. The third feature vector sequence can be a feature vector sequence obtained by combining the first feature vector sequence and the second feature vector sequence. The manner of determining the similarity between the first signal list and the second signal list based on the third feature vector sequence is similar to the manner of determining the similarity between the first signal list and the second signal list based on the second feature vector sequence, and details can be referred to the manner of determining the similarity between the first signal list and the second signal list based on the second feature vector sequence, which will not be described herein.

[0100] In S205, an object recognition result corresponding to the first signal list is output based on the similarity.

[0101] In this embodiment, the first feature vector sequence between the first signal list and different second signal lists is constructed, and at least part of the feature vectors in the first feature vector sequence is updated according to the correlation between different feature vectors in the first feature vector sequence to obtain a second feature vector sequence. The similarity between the first signal list and the second signal list is determined based on the second feature vector sequence, which can further improve the accuracy of the determined similarity between the first signal list and the second signal list, and further improve the accuracy of the generated object recognition result.

[0102] In an optional implementation, taking the identification of a visited point of interest based on a WIFI signal list by using a model as an example, after receiving a point of interest identification request (i.e., an object recognition request) sent by a user, the user equipment can provide a point of interest identification service for the user based on the wireless signal (such as a WIFI signal) scanned by the user equipment. Unlike the Global Positioning System (GPS) positioning which needs to return the latitude and longitude of the user, the point of interest visit identification needs to identify the point of interest where the user is from multiple points of interest near the user in a more complex indoor scene in most cases.

[0103] In the related art, the point of interest identification is generally based on a strategy rule. However, the point of interest identification based on the strategy rule needs to rely on feature engineering, and the online deployment cost is high.

[0104] In view of this, this embodiment provides a serialization-based modeling scheme, which implicitly encodes the input WIFI signal within the deep learning model, replacing the tedious calculations of feature engineering and greatly reducing the online deployment cost of the model; furthermore, the serialization encoding based on the self-attention mechanism enables the model to comprehensively consider different WIFI signals when characterizing fingerprint similarity, thereby improving the model's recognition accuracy.

[0105] Figure 3 A schematic diagram of the result of a model provided for an embodiment, such as Figure 3 As shown, this embodiment implements point-of-interest (POI) visit identification based on WIFI fingerprint positioning technology. Specifically, it inputs the scanned WIFI list (i.e., the first signal list) from the user's location request (i.e., object identification request), and the historical WIFI visit distribution corresponding to the POI (i.e., the second signal list). Based on model calculations, the probability of the user actually visiting this POI is obtained. The model can be constructed based on the basic structure of the BERT model. First, through a common WIFI signal encoding layer, various information of each WIFI in each list (WIFI MAC address, relationship with the POI, scanned signal strength, etc.) is encoded into corresponding encoding vectors. The encoding vectors of the request WIFI (i.e., the first signal) and the fingerprint WIFI (i.e., the second signal) are concatenated to form a two-dimensional vector (i.e., the first feature vector sequence), which is input into the attention layer for attention extraction. A residual connection layer retains some of the original WIFI signal information, and finally, dimensionality reduction is used to obtain the probability of the user actually visiting the POI (i.e., similarity).

[0106] Specifically, regarding the WIFI signal encoding layer, whether it's the actual historical distribution of WIFIs visited at points of interest or the scanned WIFI list reported by the user at the current location, the signal information of each WIFI has similar dimensions, including the following multiple dimensions of signal information:

[0107] The MAC address (identification information) of the Wi-Fi network;

[0108] Whether the WIFI is an associated WIFI of the point of interest (i.e., object association information), where associated WIFI means that the routing device corresponding to this WIFI belongs to the current point of interest, which can be obtained by mining based on the similarity between the name of the point of interest and the WIFI;

[0109] The historical signal strength distribution of WIFI visits (i.e., signal strength distribution information), where for a single report, the signal strength is 100% distributed within a certain signal strength range;

[0110] WIFI's history visit index distribution (i.e., the first ranking distribution information), which can be understood as the WIFI ranking list after the signal strength is sorted, which should be ranked in the first few places. For single reporting, the index 100% distribution is in a certain index interval (i.e., the first index interval);

[0111] WIFI's full collection signal strength distribution (i.e., the second ranking distribution information), such as the signal strength distribution of the WIFI in all requests to scan this WIFI.

[0112] Further, the features in these dimensions are vectorized (Embedding), and the Embedding results of the above five kinds of information are directly added to obtain the final WIFI signal coding output of the corresponding WIFI signal, that is, the feature vector of the WIFI signal.

[0113] For the self-attention layer, the self-attention layer is the key to sequential modeling. By introducing the self-attention mechanism, the interaction relationship between different WIFI is modeled to achieve better model effect. Unlike text modeling, the WIFI position relationship described in this embodiment is not only a linear relationship based on ranking, but also further considers the difference in WIFI signal strength. Thus, a bias term can be added after the QK matrix, and the attention calculation formula is modified as:

[0114]

[0115] Wherein,

[0116] W pos =W rssi α rssi +W index α index ,

[0117] W rssi and W index are two bias item encoding matrices that can be trained to encode the "position" relationship between different WIFI. α rssi and α index , the distribution difference between different WIFI signal strength / index, is represented by the Wasserstein distance.

[0118] Therefore, this embodiment can reduce the cost of manual feature engineering, and the modeling method can quickly migrate in different application scenarios to improve the generalization of the model. Moreover, based on the information of the global WIFI list, the accuracy of the point of interest visit recognition model can be improved.

[0119] Figure 4A structural block diagram of an object identification device is provided in the embodiments of the present disclosure. The device can be implemented by software and / or hardware, and can be configured in an electronic device, typically, a computer, a mobile phone or a tablet computer. The object identification method can be executed to identify an object based on a signal list, such as a wireless network communication WIFI signal scanned by a user when requesting a visit interest point. Figure 4 As shown in the figure, the object identification device provided in the embodiments can include a list obtaining module 401, a similarity determining module 402 and a result output module 403, wherein,

[0120] The list obtaining module 401 is configured to obtain a second signal list of at least one first historical object in response to an object identification request based on a first signal list.

[0121] The similarity determining module 402 is configured to determine a similarity between the first signal list and the second signal list according to a correlation between a first signal in the first signal list and different second signals in the second signal list.

[0122] The result output module 403 is configured to output an object identification result corresponding to the first signal list based on the similarity.

[0123] The object identification device provided in the embodiments is configured to obtain a second signal list of at least one first historical object in response to an object identification request based on a first signal list by the list obtaining module, determine a similarity between the first signal list and each second signal list according to a correlation between a first signal in the first signal list and different second signals in the second signal list by the similarity determining module, and output an object identification result corresponding to the first signal list based on the similarity by the result output module. The above technical solution is used in the embodiments to comprehensively consider the correlation between a first signal in a first signal list and different second signals in a second signal list when identifying an object based on the first signal list, so as to reduce the situation of object misidentification and improve the accuracy of the generated object identification result.

[0124] Optionally, the similarity determining module 402 can comprise: a sequence generating unit configured to generate a first feature vector sequence between the first signal list and the second signal list, the first feature vector sequence comprising feature vectors of a plurality of signals, the plurality of signals comprising at least part of the first signals in the first signal list and at least part of the second signals in the second signal list; a vector updating unit configured to update at least part of the feature vectors in the first feature vector sequence according to the correlation between different feature vectors in the first feature vector sequence, to obtain a second feature vector sequence; and a similarity determining unit configured to determine the similarity between the first signal list and the second signal list based on the second feature vector sequence.

[0125] Optionally, the vector updating unit can be specifically configured to: perform attention calculation on the feature vectors in the first feature vector sequence by using a preset attention calculation formula; and update at least part of the feature vectors in the first feature vector sequence based on the calculation result of the attention calculation.

[0126] Optionally, the bias term matrix in the preset attention calculation formula is associated with difference information between the plurality of signals in pairs, and the difference information comprises first difference information of signal intensity distribution and / or second difference information of signal intensity order distribution.

[0127] Optionally, the sequence generating unit can be specifically configured to: generate feature vectors of at least part of the first signals in the first signal list according to signal information of the at least part of the first signals; generate feature vectors of at least part of the second signals in the second signal list according to signal information of the at least part of the second signals; and construct the first feature vector sequence between the first signal list and the second signal list based on the feature vectors of the at least part of the first signals and the feature vectors of the at least part of the second signals.

[0128] Optionally, the similarity determining unit can be specifically configured to: determine the similarity between the first signal list and the second signal list based on the first feature vector sequence and the second feature vector sequence.

[0129] Optionally, the similarity determining module 402 can be specifically configured to: input the first signal list and the second signal list into a pre-trained model, and determine the similarity between the first signal list and the second signal list by the model according to the correlation between the first signals in the first signal list and different second signals in the second signal list.

[0130] Optionally, the list obtaining module 401 can be specifically configured to: obtain a second signal list of at least one first historical object according to the first signal list.

[0131] Optionally, the list obtaining module 401 can be specifically configured to: obtain at least one third signal list having a repeated signal with the first signal list from an object library, wherein the object library stores signal lists of a plurality of historical objects; obtain at least part of the third signal list from the at least one third signal list as a second signal list, and take a historical object corresponding to the second signal list as a first historical object, wherein a signal quantity of the repeated signal between the at least part of the third signal list and the first signal list satisfies a preset quantity condition.

[0132] Optionally, the result output module 403 can be configured to perform at least one of the following: in response to the presence of a second historical object in the at least one first historical object, output a first object recognition result corresponding to the first signal list based on the second historical object, wherein a similarity between a second signal list of the second historical object and the first signal list satisfies a preset similarity condition; and in response to the absence of a second historical object in the at least one first historical object, output a second object recognition result corresponding to the first signal list.

[0133] The object recognition apparatus provided by the embodiments of the present disclosure can execute the object recognition method provided by any of the embodiments of the present disclosure, and has the corresponding function modules and beneficial effects of executing the object recognition method. Technical details not described in detail in the present embodiment can be referred to the object recognition method provided by any of the embodiments of the present disclosure.

[0134] Reference is made below to Figure 5 which shows a structural schematic diagram of an electronic device (for example, a server) 500 suitable for implementing the embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure can include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablets), PMPs (portable multimedia players), vehicle-mounted terminals (for example, vehicle-mounted navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. Figure 5 The electronic device shown is only an example, and should not bring any limitation to the functions and use range of the embodiments of the present disclosure.

[0135] As Figure 5As shown, the electronic device 500 can include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 502 or loaded into a random access memory (RAM) 503 from a storage device 508. Various programs and data required for the operation of the electronic device 500 are also stored in the RAM 503. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0136] Generally, the following devices can be connected to the I / O interface 505: input devices 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 508 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 509. The communication devices 509 can allow the electronic device 500 to communicate wirelessly or wired with other devices to exchange data. Although Figure 5 The electronic device 500 is shown with various devices, but it should be understood that all of the illustrated devices are not required to implement or have the electronic device. More or less devices can alternatively be implemented.

[0137] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication devices 509, or installed from the storage devices 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-described functions defined in the methods of embodiments of the present disclosure are performed.

[0138] It should be noted that the computer-readable medium described above can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program used by or in connection with an instruction execution system, apparatus or device. In the present disclosure, the computer-readable signal medium can include a data signal propagated in baseband or propagated as a carrier wave in a propagated data signal, in which the computer-readable program code is contained. Such a propagated data signal can take many forms, including but not limited to, an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium that can send, propagate or transfer the program for use by or in connection with the instruction execution system, apparatus or device. The program code contained in the computer-readable medium can be transmitted by any suitable medium, including but not limited to, wire, cable, RF (radio frequency), etc., or any suitable combination of the above.

[0139] In some embodiments, the client, server, or both can communicate using any current known or future developed network protocol, such as HTTP (HyperText Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include local area networks ("LAN"), wide area networks ("WAN"), the Internet, and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any current known or future developed networks.

[0140] The computer-readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device, and can be accessed via the electronic device.

[0141] The computer readable medium described above carries one or more programs, when the one or more programs are executed by the electronic device, cause the electronic device to: in response to an object recognition request based on a first signal list, acquire a second signal list of at least one first historical object; determine a similarity between the first signal list and the second signal list according to a correlation between a first signal in the first signal list and a different second signal in the second signal list; and output an object recognition result corresponding to the first signal list based on the similarity.

[0142] Computer program code for carrying out operations of the present disclosure can be written in any one or more programming languages, including object oriented programming languages such as Java, Smalltalk, C++ or the like, conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0143] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present disclosure. In this regard, each block in the flow diagrams or block diagrams can represent a module, a procedure, or a part of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in a different order than that noted in the figures. For example, two blocks noted in succession can in fact be executed substantially concurrently or in the opposite order, depending on the functionality involved. It should also be noted that each block in the block diagrams and / or flow diagrams, and combinations of blocks in the block diagrams and / or flow diagrams, can be implemented by dedicated hardware-based systems that perform the specified functions or operations, or can be implemented by a combination of dedicated hardware-based systems and computer instructions.

[0144] The units described in the embodiments of the present disclosure can be implemented by software, or by hardware. In some cases, the name of the module does not constitute a limitation on the unit itself.

[0145] The functionality described herein above can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Application-specific Integrated Circuits (ASICs), Application-specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0146] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0147] According to one or more embodiments of the present disclosure, example 1 provides a method for object recognition, comprising:

[0148] In response to an object recognition request based on a first signal list, obtaining a second signal list of at least one first historical object;

[0149] According to a correlation between a first signal in the first signal list and a different second signal in the second signal list, determining a similarity between the first signal list and the second signal list;

[0150] Based on the similarity, outputting an object recognition result corresponding to the first signal list.

[0151] According to one or more embodiments of the present disclosure, example 2 provides the method according to example 1, wherein the determining the similarity between the first signal list and the second signal list according to the correlation between the first signal in the first signal list and the different second signal in the second signal list comprises:

[0152] Generating a first feature vector sequence between the first signal list and the second signal list, the first feature vector sequence including feature vectors of a plurality of signals, the plurality of signals including at least part of the first signals in the first signal list and at least part of the second signals in the second signal list;

[0153] update at least part of the feature vectors in the first feature vector sequence according to correlations between different feature vectors in the first feature vector sequence, to obtain a second feature vector sequence;

[0154] determine the similarity between the first signal list and the second signal list based on the second feature vector sequence.

[0155] According to one or more embodiments of the present disclosure, in example 3, the method according to example 2, the updating of the at least part of the feature vectors in the first feature vector sequence according to the correlations between different feature vectors in the first feature vector sequence comprises:

[0156] perform attention calculation on the feature vectors in the first feature vector sequence using a preset attention calculation formula;

[0157] update the at least part of the feature vectors in the first feature vector sequence based on the calculation result of the attention calculation.

[0158] According to one or more embodiments of the present disclosure, in example 4, the method according to example 3, the bias term matrix in the preset attention calculation formula is associated with difference information between the two of the plurality of signals, and the difference information comprises first difference information of signal intensity distribution and / or second difference information of signal intensity order distribution.

[0159] According to one or more embodiments of the present disclosure, in example 5, the method according to example 2, the first signal list comprises signal information of at least one first signal, and the second signal list comprises signal information of at least one second signal, and the generating of the first feature vector sequence between the first signal list and the second signal list comprises:

[0160] generating feature vectors of the at least part of the first signals according to the signal information of the at least part of the first signals in the first signal list, and generating feature vectors of the at least part of the second signals according to the signal information of the at least part of the second signals in the second signal list;

[0161] constructing the first feature vector sequence between the first signal list and the second signal list based on the feature vectors of the at least part of the first signals and the feature vectors of the at least part of the second signals.

[0162] According to one or more embodiments of the present disclosure, in example 6, the method according to example 2, the determination of the similarity between the first signal list and the second signal list based on the second feature vector sequence comprises:

[0163] determine the similarity between the first signal list and the second signal list based on the first feature vector sequence and the second feature vector sequence.

[0164] According to one or more embodiments of the present disclosure, in example 7, the method according to any one of examples 1-6, the determining the similarity between the first signal list and the second signal list based on the correlation between the first signal in the first signal list and the different second signal in the second signal list comprises:

[0165] inputting the first signal list and the second signal list into a pre-trained model, and determining the similarity between the first signal list and the second signal list based on the correlation between the first signal in the first signal list and the different second signal in the second signal list through the model.

[0166] According to one or more embodiments of the present disclosure, in example 8, the method according to any one of examples 1-6, the obtaining the second signal list of the at least one first historical object comprises:

[0167] obtaining the second signal list of the at least one first historical object according to the first signal list.

[0168] According to one or more embodiments of the present disclosure, in example 9, the method according to example 8, the obtaining the second signal list of the at least one first historical object according to the first signal list comprises:

[0169] obtaining at least one third signal list having a repeated signal with the first signal list from an object library, wherein the object library stores signal lists of a plurality of historical objects;

[0170] obtaining at least part of the third signal list from the at least one third signal list as the second signal list, and taking the historical object corresponding to the second signal list as the first historical object, wherein the number of signals of the repeated signal between the at least part of the third signal list and the first signal list meets a preset number condition.

[0171] According to one or more embodiments of the present disclosure, in example 10, the method according to any one of examples 1-6, the outputting the object recognition result corresponding to the first signal list based on the similarity comprises at least one of:

[0172] in response to the presence of a second historical object in the at least one first historical object, outputting a first object recognition result corresponding to the first signal list based on the second historical object, wherein the similarity between the second signal list of the second historical object and the first signal list meets a preset similarity condition;

[0173] In response to the absence of a second historical object among the at least one first historical object, the second object identification result corresponding to the first signal list is output.

[0174] According to one or more embodiments of this disclosure, Example 11 provides an electronic device, including:

[0175] One or more processors;

[0176] Memory, used to store one or more programs.

[0177] When the one or more programs are executed by the one or more processors, the one or more processors implement the object recognition method as described in any of Examples 1-10.

[0178] According to one or more embodiments of the present disclosure, Example 12 provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the object recognition method as described in any of Examples 1-10.

[0179] According to one or more embodiments of this disclosure, Example 13 provides a computer program product that, when executed by a computer, causes the computer to implement the object recognition method as described in any of Examples 1-10.

[0180] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0181] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0182] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. An object recognition method, characterized in that, include: In response to an object identification request based on a first signal list, a second signal list of at least one first historical object is obtained; The similarity between the first signal list and the second signal list is determined based on the correlation between the first signal in the first signal list and the different second signals in the second signal list; Based on the similarity, output the object recognition result corresponding to the first signal list.

2. The method according to claim 1, characterized in that, Determining the similarity between the first signal list and the second signal list based on the correlation between the first signal in the first signal list and different second signals in the second signal list includes: Generate a first feature vector sequence between the first signal list and the second signal list. The first feature vector sequence includes feature vectors of multiple signals, and the multiple signals include at least a portion of the first signals in the first signal list and at least a portion of the second signals in the second signal list. Based on the correlation between different feature vectors in the first feature vector sequence, at least some feature vectors in the first feature vector sequence are updated to obtain a second feature vector sequence. The similarity between the first signal list and the second signal list is determined based on the second feature vector sequence.

3. The method according to claim 2, characterized in that, The step of updating at least a portion of the feature vectors in the first feature vector sequence based on the correlation between different feature vectors in the first feature vector sequence includes: An attention calculation is performed on the feature vectors in the first feature vector sequence using a preset attention calculation formula. Based on the calculation results of the attention calculation, at least some feature vectors in the first feature vector sequence are updated.

4. The method according to claim 3, characterized in that, The bias term matrix in the preset attention calculation formula is associated with the pairwise difference information between the multiple signals, and the difference information includes first difference information of the signal intensity distribution and / or second difference information of the signal intensity sorting distribution.

5. The method according to claim 2, characterized in that, The first signal list includes signal information of at least one first signal, and the second signal list includes signal information of at least one second signal. Generating a first feature vector sequence between the first signal list and the second signal list includes: Based on the signal information of at least a portion of the first signals in the first signal list, a feature vector of the at least a portion of the first signals is generated; and based on the signal information of at least a portion of the second signals in the second signal list, a feature vector of the at least a portion of the second signals is generated. Based on the feature vectors of at least a portion of the first signals and the feature vectors of at least a portion of the second signals, a first feature vector sequence is constructed between the first signal list and the second signal list.

6. The method according to claim 2, characterized in that, Determining the similarity between the first signal list and the second signal list based on the second feature vector sequence includes: Based on the first feature vector sequence and the second feature vector sequence, the similarity between the first signal list and the second signal list is determined.

7. The method according to any one of claims 1-6, characterized in that, Determining the similarity between the first signal list and the second signal list based on the correlation between the first signal in the first signal list and different second signals in the second signal list includes: The first signal list and the second signal list are input into a pre-trained model. The model determines the similarity between the first signal list and the second signal list based on the correlation between the first signal in the first signal list and the different second signals in the second signal list.

8. The method according to any one of claims 1-6, characterized in that, The second signal list for obtaining at least one first historical object includes: Obtain a second signal list for at least one first historical object based on the first signal list.

9. The method according to claim 8, characterized in that, The step of obtaining a second signal list of at least one first historical object based on the first signal list includes: Obtain at least one third signal list from the object library that has a duplicate signal with the first signal list, wherein the object library stores signal lists of multiple historical objects; At least a portion of the third signal list is obtained from the at least one third signal list as a second signal list, and the historical object corresponding to the second signal list is used as a first historical object, wherein the number of repeated signals between the at least a portion of the third signal list and the first signal list satisfies a preset quantity condition.

10. The method according to any one of claims 1-6, characterized in that, The object recognition result corresponding to the first signal list is output based on the similarity, including at least one of the following: In response to the existence of a second historical object among the at least one first historical object, a first object identification result corresponding to the first signal list is output based on the second historical object, wherein the similarity between the second signal list of the two historical objects and the first signal list satisfies a preset similarity condition; In response to the absence of a second historical object among the at least one first historical object, the second object identification result corresponding to the first signal list is output.

11. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the object recognition method according to any one of claims 1-10.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the object recognition method according to any one of claims 1-10.

13. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the object recognition method according to any one of claims 1-10.

Images