WO2022247165A1 - 地理位置区域的编码方法、建立编码模型的方法及装置 - Google Patents
地理位置区域的编码方法、建立编码模型的方法及装置 Download PDFInfo
- Publication number
- WO2022247165A1 WO2022247165A1 PCT/CN2021/131177 CN2021131177W WO2022247165A1 WO 2022247165 A1 WO2022247165 A1 WO 2022247165A1 CN 2021131177 W CN2021131177 W CN 2021131177W WO 2022247165 A1 WO2022247165 A1 WO 2022247165A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- geographic location
- location area
- geographic
- coding
- sample
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 81
- 230000006870 function Effects 0.000 claims abstract description 70
- 238000009826 distribution Methods 0.000 claims abstract description 60
- 238000012545 processing Methods 0.000 claims abstract description 31
- 238000007499 fusion processing Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 54
- 230000008569 process Effects 0.000 claims description 14
- 230000004927 fusion Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 abstract description 8
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 11
- 238000004891 communication Methods 0.000 description 8
- 238000010187 selection method Methods 0.000 description 6
- 235000013410 fast food Nutrition 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000000977 initiatory effect Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000002650 habitual effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/907—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/909—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using geographical or spatial information, e.g. location
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/771—Feature selection, e.g. selecting representative features from a multi-dimensional feature space
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
Definitions
- the present disclosure relates to the field of computer application technology, in particular to big data and deep learning technology in artificial intelligence technology.
- the present disclosure provides a method, device, device, and computer storage medium for encoding a geographic location area, so as to realize reasonable encoding of the geographic location area.
- a method for establishing a coding model comprising:
- the training data includes more than one triplet, the triplet includes an anchor sample, a positive sample and a negative sample of the geographic location area;
- the encoding model is respectively executed for each sample: performing embedding processing on at least one geographical function information and at least one feature distribution information of the samples, and performing embedding processing on each vector representation obtained by embedding processing Fusion processing to obtain the encoding result of the sample;
- the training objectives of the coding model include: minimizing the distance between the coding results of the anchor samples in the triplet and the coding results of the positive samples, and maximizing the distance between the coding results of the anchor samples in the triplet and the coding results of the negative samples. distance.
- a method for encoding a geographic location area including:
- the encoding model Inputting the acquired geographic function information and feature distribution information into the encoding model, the encoding model performs embedding processing on the geographic function information and feature distribution information respectively, and performs fusion processing on the vector representations obtained through the embedding process to obtain the geographic location The encoding result of the region.
- an apparatus for establishing a coding model including:
- An acquisition unit configured to acquire training data, the training data includes more than one triplet, the triplet includes an anchor sample, a positive sample, and a negative sample of the geographic location area;
- the training unit is used to use the training data to train the encoding model; the encoding model is respectively executed for each sample: at least one geographical function information and at least one feature distribution information of the sample are respectively embedded, and the embedded processing is obtained Each vector representation of is fused to obtain the coding result of the sample;
- the training objectives of the coding model include: minimizing the distance between the coding results of the anchor samples in the triplet and the coding results of the positive samples, and maximizing the distance between the coding results of the anchor samples in the triplet and the coding results of the negative samples. distance.
- an encoding device for a geographic location area including:
- a determination unit is used to determine the geographic location area to be coded
- an acquisition unit configured to acquire at least one geographic function information and at least one feature distribution information of the geographic location area
- the encoding unit is used to input the acquired geographic function information and feature distribution information into a coding model, and the coding model performs embedding processing on the geographic function information and feature distribution information respectively, and performs fusion processing on each vector representation obtained by the embedding process, An encoding result of the geographic location area is obtained.
- an electronic device comprising:
- the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can perform the method as described above.
- a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the above method.
- a computer program product comprises a computer program which, when executed by a processor, implements the method as described above.
- the coding model provided by the present disclosure is coded based on the geographical function information and feature distribution information of the geographical location area, which can make the coding results of the geographical location areas with similar geographical functions and feature distributions more similar.
- This encoding method is more reasonable than the traditional encoding method.
- FIG. 1 is a flowchart of a method for establishing a coding model provided by an embodiment of the present disclosure
- FIG. 2 is a schematic structural diagram of a coding model provided by an embodiment of the present disclosure
- FIG. 3 is a flowchart of a coding method for a geographic location area provided by an embodiment of the present disclosure
- FIG. 4 is an example diagram of the result of applying geographic location area coding provided by an embodiment of the present disclosure
- FIG. 5 is a structural diagram of a device for establishing a coding model provided by an embodiment of the present disclosure
- FIG. 6 is a structural diagram of an encoding device for a geographic location area provided by an embodiment of the present disclosure
- FIG. 7 is a block diagram of an electronic device used to implement an embodiment of the present disclosure.
- a geographic area code is the representation of a geographic area with a code that distinguishes it from other geographic areas within a limited set of geographic areas.
- the geographic location area may be divided according to administrative divisions, such as a city, a district, a block, and so on. It can also be divided according to the preset precision and shape, for example, it can be divided into 1km ⁇ 1km grids, and each grid is regarded as a geographic location area. and many more.
- the coding of the geographic location area in the present disclosure is mainly realized based on the coding model, and thus mainly includes two stages: the stage of establishing the coding model and the stage of coding the geographic location area by using the coding model. The two stages are described below.
- Fig. 1 is a flowchart of a method for establishing a coding model provided by an embodiment of the present disclosure.
- the execution body of the method may be a device for establishing a coding model, which may be an application located on the server side, or may also be a plug-in in an application located on the server side Or a functional unit such as a software development kit (Software Development Kit, SDK), or it can also be located in a computer terminal with strong computing capabilities, which is not particularly limited in the embodiment of the present invention.
- the method may include the following steps:
- training data is obtained, and the training data includes more than one triplet, and the triplet includes an anchor sample, a positive sample, and a negative sample of a geographic location area.
- the training data to train the encoding model uses the training data to train the encoding model; the encoding model is executed for each sample: perform embedding processing on at least one geographical function information and at least one feature distribution information of the samples, and perform embedding processing on each vector representation obtained by the embedding processing Fusion processing to obtain the encoding result of the sample; the training objectives of the encoding model include: minimizing the distance between the encoding result of the anchor sample in the triplet and the encoding result of the positive sample, maximizing the encoding result of the anchor sample in the triplet and the negative The distance between encoded results of samples.
- the established coding model is coded based on the geographical function information and feature distribution information of the geographic location area, which can make the coding results of the geographic location areas with similar geographic functions and feature distribution more accurate. resemblance. The above steps will be described in detail below with reference to the embodiments.
- step 101 that is, "acquiring training data"
- step 101 that is, "acquiring training data”
- each triplet includes anchor samples, positive samples and negative samples.
- Each sample is a geographic area.
- the positive samples are geographic regions that are very similar to the anchor samples in terms of geographic function and feature distribution.
- Negative samples are geographic regions that are not similar to anchor samples in terms of geographic function and feature distribution.
- Each triple in the training data can be manually selected. Although this method has higher accuracy, it consumes more labor costs and has lower efficiency. Therefore, the embodiments of the present disclosure provide several ways to automatically acquire training data, such as but not limited to the following:
- the first method obtain the anchor samples of the geographical location area, select the neighboring geographical location area of the anchor sample as the positive sample, and select the non-neighboring geographical location area of the anchor sample as the negative sample.
- the anchor samples may be selected from the pre-divided geographical locations.
- the positive sample of the anchor sample since two adjacent geographical locations are more likely to be similar in geographical function and feature distribution, one can be selected as a positive sample from the anchor sample’s neighboring geographic location.
- the selection method may be a random selection method, or a selection method according to certain rules.
- selecting negative samples for anchor samples one can be selected from non-neighboring geolocation regions as negative samples.
- the selection method can also be a random selection method, or a selection method according to certain rules.
- the second method from the navigation log, the geographic location area where the navigation starting point is located and the geographic location area where the navigation end point is located are respectively used as anchor samples and positive samples of the geographic location area, and other geographic location areas are selected as negative samples.
- the origin and destination are likely to be similar in geographical function and feature distribution. Therefore, the navigation information of a large number of users can be obtained from the navigation log, and the geographical location area pairs composed of the navigation start point and the navigation end point can be counted, and the geographical location area pairs whose occurrence frequency or number of occurrences meet certain conditions can be used as anchor samples and positive samples. Negative samples of anchor samples can be randomly selected from other geographical regions except positive samples and anchor samples.
- the third method from the retrieval log, obtain the geographical location area where the retrieval initiation location is located and the geographic location area where the target location is located as the anchor sample and positive sample of the geographic location area respectively, and randomly select other geographic location areas as negative samples.
- the location where the user initiates the search and the target location of the search are likely to be similar in geographical function and feature distribution. Therefore, retrieval information of a large number of users can be obtained from retrieval logs.
- Statistics are made on the geographic location area pairs formed by the retrieval initiation location and the target location, and the geographic location area pairs whose occurrence frequency or number of occurrences meet certain conditions are used as anchor samples and positive samples. Negative samples of anchor samples can be randomly selected from other geographical regions except positive samples and anchor samples.
- step 102 that is, "training the encoding model by using the training data"
- At least one geographic function information and at least one feature distribution information are extracted from the samples in the triplet.
- the geographic function information may include at least one of POI (Point Of Interest, point of interest) information, user information, and location query terms initiated in the geographic location area.
- POI Point Of Interest, point of interest
- the POI information may include the POI name, POI type, POI quantity, address, etc. included in the geographic location area. These POI information can largely reflect the geographical function of the geographical location area. For example, the geographic capabilities of the geographic regions where Disney and the amusement parks are located are similar.
- the user information may include age distribution, gender ratio, occupation type distribution, education status, salary status, etc. of users in the geographic location area.
- age distribution may include age distribution, gender ratio, occupation type distribution, education status, salary status, etc. of users in the geographic location area.
- the users in the geographic location areas of science and technology parks present the characteristics that they are mostly male, between 25 and 35 years old, programmers, have a college degree or above, and have relatively high salaries.
- the location query words initiated in the geographic location area largely reflect the user preference of the geographic location area, and also reflect the geographic function of the geographic location area to a certain extent. This part of data can also be obtained from the retrieval log, and the location query words initiated in the geographical location area in the retrieval log are counted to obtain the location query words whose occurrence frequency or number of occurrences meets certain conditions.
- the feature distribution information may include at least one of a map image and a real scene image of the geographic location area. These images can be obtained from the server or database of the map application.
- the map image of the geographic location area may be an image of the geographic location area displayed on a map.
- the map image can be a satellite image or a basemap image.
- Map images include map elements of various types of areas such as land, water systems, and green spaces, as well as roads such as expressways, urban main roads, and railways, as well as attractions, hotels, schools, shopping malls, shops, office buildings, Various types of POIs such as stadiums.
- the map image well reflects the distribution of features in the geographical area.
- the real scene image refers to an image drawn or photographed based on an actual scene, such as a street view image.
- Real-world images also reflect the distribution of ground objects in geographic locations very well.
- five types of POI information, user information, location query words initiated in the geographic location area, map images, and real-scene images are extracted from the geographic location area (each sample) in the subsequent embodiments. Taking features as an example, these five types of features are represented as: X 1 , X 2 , X 3 , X 4 and X 5 .
- the above five types of features X 1 , X 2 , X 3 , X 4 , and X 5 extracted from the geographical location area are input into the encoding model, and the encoding model embedding these five types of features respectively to obtain the vector representations of various features , which is the vector representation of POI information
- Vector representation of user information A vector representation of the location query terms initiated in the geographic location area
- Vector representation of map image and the vector representation of the real-world image Then, the vector representations obtained from the embedding process are fused to obtain the coding result v of the geographic location area.
- Fig. 2 is a schematic structural diagram of an encoding model provided by an embodiment of the present disclosure.
- at least two embedding networks may be included in the encoding model, and the number of embedding networks is related to the feature type extracted from the geographic location area. The quantity is consistent.
- the encoding model includes five embedded networks, denoted as M 1 , M 2 , M 3 , M 4 and M 5 .
- neural networks such as RNN can be used.
- the embedding process performed by embedding networks M 1 and M 3 can be expressed as: M 1 (X 1 , ⁇ 1 ) and M 3 (X 3 , ⁇ 3 ), where ⁇ 1 and ⁇ 3 are embedding networks M 1 , ⁇ 3 M 3 model parameters.
- a neural network such as DNN can be used.
- the embedding process performed by the embedding network M 2 can be represented as: M 2 (X 2 , ⁇ 2 ), where ⁇ 2 is a model parameter of the embedding network M 2 .
- M 4 and M 5 since the input is image data, neural networks such as CNN can be used.
- the embedding processing performed by the embedding networks M 4 and M 5 can be expressed as: M 4 (X 4 , ⁇ 4 ) and M 5 (X 5 , ⁇ 5 ), where ⁇ 4 and ⁇ 5 are the embedding networks M 4 and ⁇ 5 respectively.
- Model parameters for M 5 .
- the vector representation output by each embedding network is sent to the fusion network for fusion processing, and the encoding result v of the geographic location area is obtained.
- the fusion process may be to splicing each vector representation, and then obtain the coding result through a fully connected mapping.
- the fusion process may also be to obtain an encoding result after performing an outer product process on each vector representation.
- Other processing methods may also be used, which will not be listed here.
- the model parameters of the fused network are denoted as ⁇ .
- v a is the anchor sample
- v + is the positive sample
- v - is the negative sample.
- the training objectives of the encoding model include: minimizing the distance between the encoding results of the anchor samples in the triplet and the encoding results of the positive samples, and maximizing the distance between the encoding results of the anchor samples and the encoding results of the negative samples in the triplet.
- loss function Assuming that the processing of each sample by the above coding model is expressed as f(), then the loss function can be defined as:
- r is the preset minimum interval, the purpose is to ensure that the distance between the encoding result of the anchor sample and the encoding result of the positive sample is equal to the distance between the encoding result of the anchor sample and the encoding result of the negative sample.
- a minimum interval r. represents the Euclidean distance.
- each iteration uses the value of the loss function to update the model parameters of the encoding model, namely ⁇ 1 , ⁇ 2 , ⁇ 3 , ⁇ 4 , ⁇ 5 and ⁇ mentioned above.
- the training end conditions such as the value of the loss function converges, the preset number of iterations is reached, and so on.
- Fig. 3 is a flow chart of a method for encoding a geographic location area provided by an embodiment of the present disclosure.
- the execution body of the method may be an encoding device for a geographic location area, and the device may be an application located on the server side, or may also be an application located on the server side Functional units such as a plug-in or SDK, or may also be located in a computer terminal with strong computing capabilities, which is not particularly limited in this embodiment of the present invention.
- the method may include the following steps:
- the geographic location area to be coded is determined.
- At least one geographic function information and at least one feature distribution information of the geographic location area are acquired.
- the acquired geographic function information and feature distribution information are input into the coding model, and the coding model performs embedding processing on the geographic function information and feature distribution information respectively, and fuses the vector representations obtained through the embedding processing to obtain the geographic location The encoding result of the region.
- step 301 that is, "determining the geographic location area to be coded" will be described in detail.
- the geographic location areas pre-divided according to the preset accuracy may be used as the geographic location areas to be encoded, and the encoding results are determined one by one. It is also possible to use one of the geographic location areas as the geographic location area to be encoded to determine the encoding result.
- the geographic location coordinates of the user are obtained, the geographic location coordinates are used as input, and the geographic location area where the geographic location coordinates are located is determined as the geographic location area to be encoded.
- the encoding model can be used in real time to determine the encoding result of the geographic location area where the input geographic location coordinates are located. It is also possible to obtain and store the coding results for each geographic location area in advance, and after obtaining the input geographic location coordinates, determine the coding result of the geographic location area where the geographic location coordinates are located by querying the stored coding results of each geographic location area.
- step 302 For the acquisition of at least one geographical function information and at least one feature distribution information of the geographic location area in step 302, reference may be made to the description in step 102 in the embodiment shown in FIG. 1 , and details are not repeated here. In addition, which features are used in the process of training the encoding model, and which features are also extracted in this step.
- each feature is encoded by each embedding network to obtain each vector representation, and then the fusion network performs fusion processing on each vector representation to obtain the encoding result of the geographic location area to be encoded. That is to say, after encoding various types of features into multi-mode information, they are mapped to a unified encoding result. For example mapping to numerically encoded results.
- the geographic location area is coded by using the method in the above embodiment, it can be applied to various application scenarios.
- the following are just a few:
- the first application scenario using the distance between the encoding results of the geographical location areas to determine similar geographical location areas.
- the user when the user needs to determine an area with a specific geographical feature, he may first select an area with this feature, and use this area as a query area. The similarity calculation is performed between the encoding result of the query area and the encoding results of other geographic locations, so as to filter out the top N geographic locations, where N is a preset positive integer. These filtered geographic areas also have the specific geographic feature. For example: For example, the user wants to find areas with residential areas and rivers, or areas where residential areas are close to rivers. As shown in Fig. 4, the user first finds out a cell A near a river, and wants to find other cells similar to this. Then the geographic location area where the cell A is located may be used as the query area.
- the second application scenario Based on the coded results of the user's geographical location, search recommendations are made to the user.
- the geographic location area where the user searches is obtained, and the coded result of the geographic location area is used as one of the input features for search recommendation. For example, when “ba" is input in the input box, along with the user's input, search words will be recommended to the user through a form such as a drop-down box. If the user is located in a certain hotel in Beijing, scenic spots such as "Badaling Great Wall” are recommended to him first. If the user is located in the science and technology park, the office buildings of technology companies such as “Baidu Building" are recommended to them first.
- the third application scenario based on the coded results of the user's geographic location, sort the search results of the user.
- the geographic location area where the user searches is obtained, and the coded result of the geographic location area is used as one of the input features to sort the search results.
- This sorting method of search results enables the recommendation to be based on the geographical function and feature distribution of the geographic location area. For example, for a restaurant search initiated in a software park in Beijing and a software park in Chengdu, although the two are geographically far apart, the coding results are highly similar due to the similarity of geographical functions and feature distribution. Based on There is also a certain similarity in the search results for restaurants, for example, they all prefer fast food.
- FIG. 5 is a structural diagram of an apparatus for establishing a coding model provided by an embodiment of the present disclosure.
- the apparatus 500 may include: an acquisition unit 501 and a training unit 502 , and may also include a division unit 503 .
- the main functions of each component unit are as follows:
- the acquiring unit 501 is configured to acquire training data, the training data includes more than one triplet, and the triplet includes an anchor sample, a positive sample, and a negative sample of a geographic location area.
- the training unit 502 is used to use the training data to train the encoding model; the encoding model is respectively executed for each sample: at least one geographical function information and at least one feature distribution information of the sample are respectively embedded, and each vector obtained by the embedding process is Indicates that fusion processing is performed to obtain the encoding result of the sample.
- the training objectives of the encoding model include: minimizing the distance between the encoding results of the anchor samples in the triplet and the encoding results of the positive samples, and maximizing the distance between the encoding results of the anchor samples and the encoding results of the negative samples in the triplet.
- the geographic function information may include at least one of point of interest information, user information, and location query terms initiated in the geographic location area;
- the feature distribution information may include at least one of a map image and a real scene image.
- the acquisition unit 501 can acquire training data in the following ways, but not limited to:
- the first method obtain the anchor samples of the geographical location area, select the neighboring geographical location area of the anchor sample as a positive sample, and select the non-neighboring geographical location area of the anchor sample as a negative sample.
- the second method from the navigation log, the geographic location area where the navigation starting point is located and the geographic location area where the navigation end point is located are respectively used as anchor samples and positive samples of the geographic location area, and other geographic location areas are selected as negative samples.
- the third method from the retrieval log, the geographical location area where the retrieval initiation location is located and the geographic location area where the target location is located are respectively used as anchor samples and positive samples of the geographical location area, and other geographical location areas are selected as negative samples.
- the division unit 503 is configured to pre-divide geographic location areas according to a preset accuracy.
- the encoding model may include: at least two embedding networks and a fusion network.
- the training unit 502 may input at least one geographical function information and at least one feature distribution information extracted from the samples into each embedding network respectively.
- the embedding network is used to embed the input information to obtain the corresponding vector representation.
- the fusion network is used to fuse the vector representations output by each embedding network to obtain the coding result of the sample.
- the training unit 502 When training the coding model, the training unit 502 iteratively updates the model parameters of the embedding network and the fusion network according to the value of the loss function, wherein the loss function is pre-constructed according to the training target.
- FIG. 6 is a structural diagram of a coding device for a geographic location area provided by an embodiment of the present disclosure. As shown in FIG. application unit 605 .
- the determining unit 601 is configured to determine the geographic location area to be coded.
- the obtaining unit 602 is configured to obtain at least one kind of geographical function information and at least one kind of feature distribution information of the geographic location area.
- the encoding unit 603 is used to input the acquired geographical function information and feature distribution information into the encoding model, and the encoding model performs embedding processing on the geographical function information and feature distribution information respectively, and fuses the vector representations obtained through the embedding process to obtain The encoded result of the geographic area.
- the geographic function information includes at least one of point of interest information, user information, and location query words initiated in the geographic location area;
- the feature distribution information includes at least one of a base map image and a street view image.
- the division unit 604 is configured to pre-divide geographic location areas according to a preset accuracy.
- the determining unit 601 acquires the input geographic location coordinates; and determines the geographic location area where the geographic location coordinates are located as the geographic location area to be coded.
- the determining unit 601 may use the divided geographic location areas as geographic location areas to be coded respectively.
- the application unit 605 is configured to use the distance between the coding results of the geographical location areas to determine similar geographical location areas; or, based on the coding results of the geographical location areas where the user is located, to perform search recommendations or sort the search results for the user.
- each embodiment in this specification is described in a progressive manner, the same and similar parts of each embodiment can be referred to each other, and each embodiment focuses on the differences from other embodiments.
- the description is relatively simple, and for relevant parts, please refer to part of the description of the method embodiment.
- the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.
- FIG. 7 it is a block diagram of an electronic device according to a method for encoding a geographic location area and a method for establishing an encoding model according to an embodiment of the present disclosure.
- Electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers.
- Electronic devices may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices.
- the components shown herein, their connections and relationships, and their functions, are by way of example only, and are not intended to limit implementations of the disclosure described and/or claimed herein.
- the device 700 includes a computing unit 701 that can execute according to a computer program stored in a read-only memory (ROM) 702 or loaded from a storage unit 708 into a random-access memory (RAM) 703. Various appropriate actions and treatments. In the RAM 703, various programs and data necessary for the operation of the device 700 can also be stored.
- the computing unit 701, ROM 702, and RAM 703 are connected to each other through a bus 704.
- An input/output (I/O) interface 705 is also connected to the bus 704 .
- the I/O interface 705 includes: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a magnetic disk, an optical disk, etc. ; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, and the like.
- the communication unit 709 allows the device 700 to exchange information/data with other devices over a computer network such as the Internet and/or various telecommunication networks.
- the computing unit 701 may be various general-purpose and/or special-purpose processing components having processing and computing capabilities. Some examples of computing units 701 include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processing processor (DSP), and any suitable processor, controller, microcontroller, etc.
- the calculation unit 701 executes various methods and processes described above, such as the method of encoding geographic location areas and the method of establishing encoding models. For example, in some embodiments, the method of encoding a geographic location area, the method of building an encoding model may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 708 .
- part or all of the computer program may be loaded and/or installed on the device 700 via the ROM 802 and/or the communication unit 709.
- the computer program When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the above-described coding method for the geographical location area and the method for establishing the coding model can be executed.
- the computing unit 701 may be configured in any other appropriate way (for example, by means of firmware) to execute a method for encoding a geographic location area and a method for establishing an encoding model.
- Various implementations of the systems and techniques described herein can be implemented in digital electronic circuitry, systems integrated circuits, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips system (SOC), complex programmable logic device (CPLD), computer hardware, firmware, software, and/or a combination thereof.
- FPGAs field programmable gate arrays
- ASICs application specific integrated circuits
- ASSPs application specific standard products
- SOC systems on chips system
- CPLD complex programmable logic device
- computer hardware firmware, software, and/or a combination thereof.
- programmable processor can be special-purpose or general-purpose programmable processor, can receive data and instruction from storage system, at least one input device, and at least one output device, and transmit data and instruction to this storage system, this at least one input device, and this at least one output device an output device.
- Program codes for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus, so that the program codes, when executed by the processor or controller, make the flow diagrams and/or block diagrams specified The function/operation is implemented.
- the program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.
- a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device.
- a machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium.
- a machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing.
- machine-readable storage media would include one or more wire-based electrical connections, portable computer discs, hard drives, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, compact disk read only memory (CD-ROM), optical storage, magnetic storage, or any suitable combination of the foregoing.
- RAM random access memory
- ROM read only memory
- EPROM or flash memory erasable programmable read only memory
- CD-ROM compact disk read only memory
- magnetic storage or any suitable combination of the foregoing.
- the systems and techniques described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user. ); and a keyboard and pointing device (eg, a mouse or a trackball) through which a user can provide input to the computer.
- a display device e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor
- a keyboard and pointing device eg, a mouse or a trackball
- Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and can be in any form (including Acoustic input, speech input or, tactile input) to receive input from the user.
- the systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., as a a user computer having a graphical user interface or web browser through which a user can interact with embodiments of the systems and techniques described herein), or including such backend components, middleware components, Or any combination of front-end components in a computing system.
- the components of the system can be interconnected by any form or medium of digital data communication, eg, a communication network. Examples of communication networks include: Local Area Network (LAN), Wide Area Network (WAN) and the Internet.
- a computer system may include clients and servers.
- Clients and servers are generally remote from each other and typically interact through a communication network.
- the relationship of client and server arises by computer programs running on the respective computers and having a client-server relationship to each other.
- the server can be a cloud server, also known as cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the management problems existing in traditional physical host and virtual private server (VPs, VI irtual Private Server) services. Difficulty and weak business expansion.
- the server can also be a server of a distributed system, or a server combined with a blockchain.
- steps may be reordered, added or deleted using the various forms of flow shown above.
- each step described in the present application may be executed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in the present disclosure can be achieved, no limitation is imposed herein.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Multimedia (AREA)
- Remote Sensing (AREA)
- Library & Information Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
Claims (21)
- 一种建立编码模型的方法,包括:获取训练数据,所述训练数据包括一个以上的三元组,所述三元组包括地理位置区域的锚样本、正样本和负样本;利用所述训练数据训练编码模型;所述编码模型针对各样本分别执行:对样本的至少一种地理功能信息和至少一种地物分布信息分别进行嵌入处理,将嵌入处理得到的各向量表示进行融合处理,得到样本的编码结果;所述编码模型的训练目标包括:最小化三元组中锚样本的编码结果和正样本的编码结果之间的距离,最大化三元组中锚样本的编码结果和负样本的编码结果之间的距离。
- 根据权利要求1所述的方法,其中,所述地理功能信息包括兴趣点信息、用户信息和在该地理位置区域发起的地点查询词中的至少一种;所述地物分布信息包括地图图像和实景图像中的至少一种。
- 根据权利要求1所述的方法,其中,所述获取训练数据包括:获取地理位置区域的锚样本,选取锚样本的邻居地理位置区域作为正样本,选取锚样本的非邻居地理位置区域作为负样本;或者,从导航日志中,获取导航起始点所在的地理位置区域和导航终点所在的地理位置区域分别作为地理位置区域的锚样本和正样本,选取其他地理位置区域作为负样本;或者,从检索日志中,获取检索的发起位置所在的地理位置区域和目标位置所在的地理位置区域分别作为地理位置区域的锚样本和正样本,选取其他地理位置区域作为负样本。
- 根据权利要求1至3中任一项所述的方法,还包括:按照预设的精度预先划分地理位置区域。
- 根据权利要求1至3中任一项所述的方法,其中,所述编码模型包括:至少两个嵌入网络以及融合网络;将从样本中提取的至少一种地理功能信息和至少一种地物分布信息分别输入各嵌入网络;所述嵌入网络对输入的信息进行嵌入处理,得到对应的向量表示;所述融合网络对各嵌入网络输出的向量表示进行融合处理,得到样本的编码结果;在训练所述编码模型时,依据损失函数的取值迭代更新所述嵌入网络和融合网络的模型参数,其中,所述损失函数依据所述训练目标预先构建得到。
- 一种地理位置区域的编码方法,包括:确定待编码的地理位置区域;获取所述地理位置区域的至少一种地理功能信息和至少一种地物分布信息;将获取的地理功能信息和地物分布信息输入编码模型,所述编码模型对地理功能信息和地物分布信息分别进行嵌入处理,将嵌入处理得到的各向量表示进行融合处理,得到所述地理位置区域的编码结果。
- 根据权利要求6所述的方法,其中,所述地理功能信息包括兴趣点信息、用户信息和在该地理位置区域发起的地点查询词中的至少一种;所述地物分布信息包括底图图像和街景图像中的至少一种。
- 根据权利要求6所述的方法,还包括:按照预设的精度预先划分地理位置区域;所述确定待编码的地理位置区域包括:获取输入的地理位置坐标;确定所述地理位置坐标所在的地理位置区域作为所述待编码的地理位置区域。
- 根据权利要求6至8中任一项所述的方法,还包括:利用地理位置区域的编码结果之间的距离,确定相似的地理位置区域;或者,基于用户所在地理位置区域的编码结果,对所述用户进行搜索推荐或搜索结果排序。
- 一种建立编码模型的装置,包括:获取单元,用于获取训练数据,所述训练数据包括一个以上的三元组,所述三元组包括地理位置区域的锚样本、正样本和负样本;训练单元,用于利用所述训练数据训练编码模型;所述编码模型针对各样本分别执行:对样本的至少一种地理功能信息和至少一种地物分布信息分别进行嵌入处理,将嵌入处理得到的各向量表示进行融合处理, 得到样本的编码结果;所述编码模型的训练目标包括:最小化三元组中锚样本的编码结果和正样本的编码结果之间的距离,最大化三元组中锚样本的编码结果和负样本的编码结果之间的距离。
- 根据权利要求10所述的装置,其中,所述地理功能信息包括兴趣点信息、用户信息和在该地理位置区域发起的地点查询词中的至少一种;所述地物分布信息包括地图图像和实景图像中的至少一种。
- 根据权利要求10所述的装置,其中,所述获取单元具体用于:获取地理位置区域的锚样本,选取锚样本的邻居地理位置区域作为正样本,选取锚样本的非邻居地理位置区域作为负样本;或者,从导航日志中,获取导航起始点所在的地理位置区域和导航终点所在的地理位置区域分别作为地理位置区域的锚样本和正样本,选取其他地理位置区域作为负样本;或者,从检索日志中,获取检索的发起位置所在的地理位置区域和目标位置所在的地理位置区域分别作为地理位置区域的锚样本和正样本,选取其他地理位置区域作为负样本。
- 根据权利要求10至12中任一项所述的装置,还包括:划分单元,用于按照预设的精度预先划分地理位置区域。
- 根据权利要求10至12中任一项所述的装置,其中,所述编码模型包括:至少两个嵌入网络以及融合网络;所述训练单元,具体用于将从样本中提取的至少一种地理功能信息和至少一种地物分布信息分别输入各嵌入网络;所述嵌入网络,用于对输入的信息进行嵌入处理,得到对应的向量表示;所述融合网络,用于对各嵌入网络输出的向量表示进行融合处理,得到样本的编码结果;所述训练单元在训练所述编码模型时,依据损失函数的取值迭代更新所述嵌入网络和融合网络的模型参数,其中,所述损失函数依据所述训练目标预先构建得到。
- 一种地理位置区域的编码装置,包括:确定单元,用于确定待编码的地理位置区域;获取单元,用于获取所述地理位置区域的至少一种地理功能信息和至少一种地物分布信息;编码单元,用于将获取的地理功能信息和地物分布信息输入编码模型,所述编码模型对地理功能信息和地物分布信息分别进行嵌入处理,将嵌入处理得到的各向量表示进行融合处理,得到所述地理位置区域的编码结果。
- 根据权利要求15所述的装置,其中,所述地理功能信息包括兴趣点信息、用户信息和在该地理位置区域发起的地点查询词中的至少一种;所述地物分布信息包括底图图像和街景图像中的至少一种。
- 根据权利要求15所述的装置,还包括:划分单元,用于按照预设的精度预先划分地理位置区域;所述确定单元,具体用于获取输入的地理位置坐标;确定所述地理位置坐标所在的地理位置区域作为所述待编码的地理位置区域。
- 根据权利要求15至17中任一项所述的装置,还包括:应用单元,用于利用地理位置区域的编码结果之间的距离,确定相似的地理位置区域;或者,基于用户所在地理位置区域的编码结果,对所述用户进行搜索推荐或搜索结果排序。
- 一种电子设备,包括:至少一个处理器;以及与所述至少一个处理器通信连接的存储器;其中,所述存储器存储有可被所述至少一个处理器执行的指令,所述指令被所述至少一个处理器执行,以使所述至少一个处理器能够执行权利要求1-9中任一项所述的方法。
- 一种存储有计算机指令的非瞬时计算机可读存储介质,其中,所述计算机指令用于使所述计算机执行权利要求1-9中任一项所述的方法。
- 一种计算机程序产品,包括计算机程序,所述计算机程序在被处理器执行时实现根据权利要求1-9中任一项所述的方法。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP21918110.4A EP4174712A4 (en) | 2021-05-24 | 2021-11-17 | CODING METHOD AND DEVICE FOR A GEOGRAPHICAL LOCATION AREA AND METHOD AND DEVICE FOR PRODUCING A CODING MODEL |
KR1020227024676A KR20220160534A (ko) | 2021-05-24 | 2021-11-17 | 지리적 위치 구역의 인코딩 방법, 인코딩 모델을 구축하는 방법 및 장치 |
US17/793,999 US20240177469A1 (en) | 2021-05-24 | 2021-11-17 | Method and apparatus for encoding geographic location region as well as method and apparatus for establishing encoding model |
JP2022543043A JP2023530795A (ja) | 2021-05-24 | 2021-11-17 | 地理的位置区域の符号化方法、符号化モデルを確立する方法、及び装置 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110565434.6 | 2021-05-24 | ||
CN202110565434.6A CN113342912B (zh) | 2021-05-24 | 2021-05-24 | 地理位置区域的编码方法、建立编码模型的方法及装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022247165A1 true WO2022247165A1 (zh) | 2022-12-01 |
Family
ID=77471102
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2021/131177 WO2022247165A1 (zh) | 2021-05-24 | 2021-11-17 | 地理位置区域的编码方法、建立编码模型的方法及装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US20240177469A1 (zh) |
EP (1) | EP4174712A4 (zh) |
JP (1) | JP2023530795A (zh) |
KR (1) | KR20220160534A (zh) |
CN (1) | CN113342912B (zh) |
WO (1) | WO2022247165A1 (zh) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113342912B (zh) * | 2021-05-24 | 2022-03-18 | 北京百度网讯科技有限公司 | 地理位置区域的编码方法、建立编码模型的方法及装置 |
CN113837162B (zh) * | 2021-11-29 | 2022-04-08 | 腾讯科技(深圳)有限公司 | 数据处理方法及相关装置 |
CN114926655B (zh) * | 2022-05-20 | 2023-09-26 | 北京百度网讯科技有限公司 | 地理与视觉跨模态预训练模型的训练方法、位置确定方法 |
CN114998684B (zh) * | 2022-05-20 | 2023-06-23 | 北京百度网讯科技有限公司 | 地理与视觉跨模态预训练模型的训练方法、定位调整方法 |
KR102664445B1 (ko) * | 2023-01-12 | 2024-05-10 | 이다경 | 건축토지를 매칭하고 실내 인테리어를 제공하는 서비스 제공 장치, 시스템, 방법 및 프로그램 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110727816A (zh) * | 2018-06-29 | 2020-01-24 | 百度在线网络技术(北京)有限公司 | 兴趣点类别确定方法和装置 |
CN110866140A (zh) * | 2019-11-26 | 2020-03-06 | 腾讯科技(深圳)有限公司 | 图像特征提取模型训练方法、图像搜索方法及计算机设备 |
US20200257940A1 (en) * | 2019-02-12 | 2020-08-13 | Canon Kabushiki Kaisha | Method, system and apparatus for generating training samples for matching objects in a sequence of images |
CN111666462A (zh) * | 2020-04-28 | 2020-09-15 | 百度在线网络技术(北京)有限公司 | 地理位置的推荐方法、装置、设备和计算机存储介质 |
CN113342912A (zh) * | 2021-05-24 | 2021-09-03 | 北京百度网讯科技有限公司 | 地理位置区域的编码方法、建立编码模型的方法及装置 |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103609144A (zh) * | 2011-06-16 | 2014-02-26 | 诺基亚公司 | 用于解析地理标识的方法和装置 |
CN108628904B (zh) * | 2017-03-23 | 2021-03-02 | 北京嘀嘀无限科技发展有限公司 | 一种路径编码、相似路径检索方法及装置和电子设备 |
CN110008293B (zh) * | 2019-02-02 | 2022-12-27 | 创新先进技术有限公司 | 地理位置查询方法及装置 |
US11948472B2 (en) * | 2019-07-30 | 2024-04-02 | Ntt Docomo, Inc. | Popularity evaluation system and geographical feature generation model |
CN111274936B (zh) * | 2020-01-19 | 2023-04-18 | 中国科学院上海高等研究院 | 多光谱图像地物分类方法、系统、介质及终端 |
CN111666292B (zh) * | 2020-04-24 | 2023-05-26 | 百度在线网络技术(北京)有限公司 | 用于检索地理位置的相似度模型建立方法和装置 |
CN111666461B (zh) * | 2020-04-24 | 2023-05-26 | 百度在线网络技术(北京)有限公司 | 检索地理位置的方法、装置、设备和计算机存储介质 |
CN112269925B (zh) * | 2020-10-19 | 2024-03-22 | 北京百度网讯科技有限公司 | 一种获取地理位置点信息的方法和装置 |
CN112559885B (zh) * | 2020-12-25 | 2024-01-12 | 北京百度网讯科技有限公司 | 地图兴趣点的训练模型确定方法、装置及电子设备 |
-
2021
- 2021-05-24 CN CN202110565434.6A patent/CN113342912B/zh active Active
- 2021-11-17 EP EP21918110.4A patent/EP4174712A4/en not_active Withdrawn
- 2021-11-17 JP JP2022543043A patent/JP2023530795A/ja active Pending
- 2021-11-17 KR KR1020227024676A patent/KR20220160534A/ko not_active Application Discontinuation
- 2021-11-17 US US17/793,999 patent/US20240177469A1/en active Pending
- 2021-11-17 WO PCT/CN2021/131177 patent/WO2022247165A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110727816A (zh) * | 2018-06-29 | 2020-01-24 | 百度在线网络技术(北京)有限公司 | 兴趣点类别确定方法和装置 |
US20200257940A1 (en) * | 2019-02-12 | 2020-08-13 | Canon Kabushiki Kaisha | Method, system and apparatus for generating training samples for matching objects in a sequence of images |
CN110866140A (zh) * | 2019-11-26 | 2020-03-06 | 腾讯科技(深圳)有限公司 | 图像特征提取模型训练方法、图像搜索方法及计算机设备 |
CN111666462A (zh) * | 2020-04-28 | 2020-09-15 | 百度在线网络技术(北京)有限公司 | 地理位置的推荐方法、装置、设备和计算机存储介质 |
CN113342912A (zh) * | 2021-05-24 | 2021-09-03 | 北京百度网讯科技有限公司 | 地理位置区域的编码方法、建立编码模型的方法及装置 |
Non-Patent Citations (1)
Title |
---|
See also references of EP4174712A4 |
Also Published As
Publication number | Publication date |
---|---|
EP4174712A4 (en) | 2024-03-20 |
JP2023530795A (ja) | 2023-07-20 |
EP4174712A1 (en) | 2023-05-03 |
US20240177469A1 (en) | 2024-05-30 |
KR20220160534A (ko) | 2022-12-06 |
CN113342912B (zh) | 2022-03-18 |
CN113342912A (zh) | 2021-09-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022247165A1 (zh) | 地理位置区域的编码方法、建立编码模型的方法及装置 | |
CN110726418B (zh) | 兴趣点区域确定方法、装置、设备及存储介质 | |
CN109478184B (zh) | 识别、处理和显示数据点聚类 | |
US9501577B2 (en) | Recommending points of interests in a region | |
WO2022213580A1 (zh) | 地图的生成方法、装置、电子设备及存储介质 | |
CN111597170B (zh) | 一种无损地由bim模型构建空间语义数据库的方法 | |
US8949196B2 (en) | Systems and methods for matching similar geographic objects | |
CN106462624A (zh) | 基于图块的地理编码器 | |
US9811539B2 (en) | Hierarchical spatial clustering of photographs | |
CN110309433B (zh) | 一种数据处理方法、装置及服务器 | |
CN114049462B (zh) | 一种三维模型单体化方法及装置 | |
US11893073B2 (en) | Method and apparatus for displaying map points of interest, and electronic device | |
CN107391516B (zh) | 公交站点聚合方法及装置 | |
CN115100643A (zh) | 融合三维场景语义的单目视觉定位增强方法和设备 | |
WO2020114273A1 (zh) | 商户查找方法、装置、电子设备和存储介质 | |
She et al. | 3D building model simplification method considering both model mesh and building structure | |
WO2015192716A1 (zh) | 一种基于电子地图的划线搜索方法和装置 | |
WO2023185144A1 (zh) | 基于geohash的空间数据处理方法、装置及电子设备 | |
US11983204B1 (en) | Mobile terminal-oriented complex condition geographic information query method, device and medium | |
US20220237474A1 (en) | Method and apparatus for semanticization, electronic device and readable storage medium | |
Caruso | A Distributed System For Large-Scale Geolocalization Of Internet Hosts | |
CN114596385A (zh) | 一种基于无序界址点自动排序的宗地图绘制方法 | |
CN117195367A (zh) | 一种异形建筑结构展现方法、装置、电子设备及存储介质 | |
CN117195149A (zh) | 一种多源数据融合方法、装置、电子设备及存储介质 | |
CN118524087A (zh) | 基于ip地址的定位方法、装置、计算机设备和存储介质 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2022543043 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 17793999 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21918110 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2021918110 Country of ref document: EP Effective date: 20230127 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |