CN112070149A - Automatic target detection platform - Google Patents
Automatic target detection platform Download PDFInfo
- Publication number
- CN112070149A CN112070149A CN202010924172.3A CN202010924172A CN112070149A CN 112070149 A CN112070149 A CN 112070149A CN 202010924172 A CN202010924172 A CN 202010924172A CN 112070149 A CN112070149 A CN 112070149A
- Authority
- CN
- China
- Prior art keywords
- model
- target detection
- module
- data set
- detection platform
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 87
- 238000012549 training Methods 0.000 claims abstract description 24
- 238000006243 chemical reaction Methods 0.000 claims description 11
- 230000000694 effects Effects 0.000 claims description 11
- 238000002372 labelling Methods 0.000 claims description 4
- 230000001502 supplementing effect Effects 0.000 claims description 2
- 238000013135 deep learning Methods 0.000 abstract description 11
- 238000007726 management method Methods 0.000 description 16
- 238000012795 verification Methods 0.000 description 13
- 238000000034 method Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 230000006835 compression Effects 0.000 description 4
- 238000007906 compression Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 230000006837 decompression Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
- G06V10/95—Hardware or software architectures specially adapted for image or video understanding structured as a network, e.g. client-server architectures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Evolutionary Biology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an automatic target detection platform, which comprises: the data set management module is used for creating and managing a data set; the model training module is used for training a model to be trained according to the data set so as to generate a first target detection model; the model checking module is used for carrying out online checking on the first target detection model so as to generate a second target detection model; and the model issuing module is used for issuing the second target detection model. According to the automatic target detection platform, the technical threshold of a deep learning target detection algorithm can be greatly reduced, and the entrance difficulty of the automatic target detection platform is reduced.
Description
Technical Field
The invention relates to the technical field of computer vision, in particular to an automatic target detection platform.
Background
In recent years, the development of deep learning is gradually mature and widely applied, and especially, a target detection algorithm based on video identification makes great progress in the development background of deep learning in the last years, and the detection performance is obviously improved.
Object detection is an important task in the field of computer vision, where multiple objects in a picture can be identified and different objects can be located, giving a bounding box. The target detection is applied to the mature situation, for example, in the unmanned application, the targets such as vehicles, pedestrians, traffic lights and the like can be detected; in the application of security monitoring, the human face, the human shape, the object and the like can be detected.
In the related art, although the target detection algorithm is well applied, according to the characteristics of supervised learning in deep learning, the development process of one target detection algorithm is not simple and can be completed by a deep learning algorithm engineer with abundant experience, so that the target detection algorithm has strong difficulty in entering a door and high technical threshold for a common user, and the target detection algorithm is limited to be popularized in more scenes of the young people.
Disclosure of Invention
The invention provides an automatic target detection platform for solving the technical problems, which can greatly reduce the technical threshold of a deep learning target detection algorithm and reduce the entrance difficulty of the deep learning target detection algorithm.
The technical scheme adopted by the invention is as follows:
an automated target detection platform, comprising: the data set management module is used for creating and managing a data set; the model training module is used for training a model to be trained according to the data set so as to generate a first target detection model; the model checking module is used for carrying out online checking on the first target detection model so as to generate a second target detection model; and the model issuing module is used for issuing the second target detection model.
The data set management module is specifically configured to: labeling the target image to obtain a sample image, performing format conversion on the sample image, and compressing the sample image after format conversion to generate the data set.
The data set management module is further specifically configured to: view the data set, and/or modify the name of the data set, and/or supplement data.
The model publishing module is specifically configured to: and issuing the second target detection model by adopting cloud service and/or offline service.
The automatic target detection platform further comprises: the identity authentication module is used for receiving identity information input by a user, authenticating the identity of the user according to the identity information, and opening the use permission of the user after the identity of the user is successfully authenticated.
The automatic target detection platform further comprises: and the model creating and inquiring module is used for creating the model to be trained and inquiring the model information of the second target detection model.
The model information includes model version number, model type, release status and model effect.
The invention has the beneficial effects that:
the invention can greatly reduce the technical threshold of the deep learning target detection algorithm and reduce the entrance difficulty of the deep learning target detection algorithm.
Drawings
FIG. 1 is a block diagram of an automated target detection platform according to an embodiment of the present invention;
FIG. 2 is a block diagram of an automated target detection platform according to one embodiment of the present invention;
FIG. 3 is a block diagram of an automated target detection platform according to another embodiment of the present invention;
FIG. 4 is a logic diagram of a method for model training based on an automated target detection platform according to an embodiment of the present invention;
FIG. 5 is a diagram of the underlying architecture of an automated target detection platform, in accordance with one embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
FIG. 1 is a block diagram of an automated target detection platform according to an embodiment of the present invention.
As shown in FIG. 1, an automated target detection platform 10 according to an embodiment of the present invention may include: data set management module 100, model training module 200, model verification module 300, and model publishing module 400.
Wherein, the data set management module 100 is used for creating and managing data sets; the model training module 200 is configured to train a model to be trained according to a data set to generate a first target detection model; the model checking module 300 is configured to perform online checking on the first target detection model to generate a second target detection model; the model publishing module 400 is configured to publish the second object detection model.
According to an embodiment of the present invention, as shown in fig. 2, the automated target detection platform 10 further comprises: the authentication module 500, wherein the authentication module 500 is configured to receive identity information input by a user, authenticate the identity of the user according to the identity information, and open the usage right of the user after the identity of the user is successfully authenticated.
Specifically, when the user uses the automated target detection platform 10, the user first needs to fill in information such as a user name, an account number, a password, and the like to complete account registration, and then enters identity information through the identity verification module 500, where the identity information includes the account number and the password. At this time, the identity authentication module 500 may authenticate the identity of the user according to the input identity information, and if the input account corresponds to the password, it indicates that the user identity authentication is successful, and at this time, the automatic target detection platform 10 may be entered and a home page may be skipped.
Further, after the user successfully logs in the automatic target detection platform 10, the user may perform model training through the automatic target detection platform.
It should be noted that, before the model is trained, the data set management module 100 may create a data set, and correspondingly, according to an embodiment of the present invention, the data set management module 100 is specifically configured to: labeling the target image to obtain a sample image, performing format conversion on the sample image, and compressing the sample image after format conversion to generate the data set.
Specifically, the creation of the data set by the data set management module 100 may include the steps of: (1) designing a label: the number of the targets to be detected can be determined according to actual requirements, the number of the tags corresponds to the types of the targets one by one, wherein the upper limit of the number of the tags is 1000, each type of target at least comprises 20 targets for ensuring the effectiveness of a data set, and in general, the number of the targets contained in each picture does not exceed 100 targets for ensuring the performance of a model; (2) ensure that the sample image meets the format requirements: converting the type of the sample image into jpg, png, bmp, jpeg, limiting the size of the sample image to be within 4M, and limiting the aspect ratio to be 3: 1, wherein the longest edge of the sample image needs to be less than 4096px, and the shortest edge of the sample image needs to be more than 20 px; (3) ensuring that the content of the sample image after format conversion meets the requirement, wherein in order to ensure the validity of the data set, before compressing the sample image after format conversion, the sample image after format conversion needs to be screened to ensure that the content of the sample image after format conversion meets the requirement, wherein the picture of the training set needs to be consistent with the picture environment to be identified in the actual scene, and the pictures of each label need to cover the possibility in the actual scene, such as the change of the photographing angle and the light brightness, the more scenes the training set covers, the stronger the generalization capability of the model; (4) the compression packet format requires: compressing the sample image after format conversion, ensuring that a compression packet is in a zip format, the size of the compression packet is within 5GB, and the compression packet comprises a picture source file and an xml file with label information, wherein the picture name is the same as the name of the xml file containing the label information and is in an inclusion relationship; (5) generating a data set: inputting a data set name, selecting a data set labeling state, adding a compressed package file, and analyzing the data set after the addition is completed, wherein the state after the data set is uploaded is divided into two states of decompression and analysis, and related information of the data set can be displayed in a data list.
According to an embodiment of the present invention, as shown in fig. 3, the automated target detection platform 10 further comprises: and a model creating and querying module 600, wherein the model creating and querying module 600 is configured to create a model to be trained and query model information of the second target detection model.
According to one embodiment of the invention, the model information includes a model version number, a model type, a release status, and a model effect. Wherein, in the model effect, the overall situation description of the model training can be seen, which comprises the following steps: basic conclusion, mAP, F1-score, accuracy, recall. The index of the model effect is based on the training data set to randomly extract partial data, and the partial data does not participate in training and only participates in the evaluation calculation of the model effect. Therefore, when the amount of data is small (for example, the number of pictures is less than 100), the data participating in the evaluation is also small, and the effect of the model evaluation report obtained in this way is only used for reference, and the effect of the model cannot be fully and accurately reflected. That is, the model effect is only used for reference, and if the model effect condition is to be known more fully, the more accurate model effect can be obtained by calling the interface batch test after the model is released as the API.
Specifically, after the data set is created, as shown in FIG. 4, the model to be trained may be created by the model creation and query module 600. The model to be trained can be created by filling in a model name, a contact way, function description information and the like through the model creating and querying module 600, and after the creation of the model to be trained is completed, the created model to be trained can be queried in the model list.
After the model to be trained is created, a data set (an existing data set or a created data set) may be added, and the model to be trained is trained by the model training module 200 according to the data set to generate a first target detection model. When the model is trained, the version which needs to be trained can be selected from the model list according to requirements. The application type selection can select cloud service (corresponding to a general model and a high-precision model) and offline service (corresponding to a general model and a high-performance model), different types correspond to different algorithms, and generally, the high-precision model is better in recognition accuracy but poorer in training speed and recognition speed; the high-performance model is better in training speed and recognition speed, but is poorer in recognition accuracy; the generic model is between the high-precision model and the high-performance model. After training is started, the user can jump to a version list (the model creating and querying module 600) under the corresponding model to check the training state, and the training can also be stopped at any time during the training process.
Further, after the trained model is trained to generate the first target detection model, the model verification module 300 may perform online verification on the first target detection model. Specifically, the model verification module mainly verifies the performance of the trained model, wherein the time for starting the model verification is related to the model, and the model verification can be started successfully after waiting for about 1-5 minutes. In the verification process, corresponding pictures may be added, and a preset number (for example, 10) of pictures may be selected for verification, so as to generate a verified object detection model, that is, a second object detection model. Wherein, the more pictures selected by the verification model, the longer the time is needed.
After verification is completed, the second target detection model may be published through the model publishing module 400, where the second target detection model may be published using cloud services and/or offline services.
In particular, the second target detection model may be deployed in a cloud service, an offline service. As a possible implementation manner, the second target detection model can be packaged into an SDK adapted to the intelligent hardware, and the device-side offline calculation can be performed; as another possible implementation manner, cloud service publishing may be adopted, wherein a service name, an interface address, or other requirements may be added to the cloud service publishing, after the cloud service submits an application, the cloud service jumps to a version list under a corresponding model to check a publishing state, and when the publishing state is "published", the model may be checked for "details of the service" or "offline"; of course, as another possible implementation manner, the cloud service and the offline service may be published simultaneously.
According to an embodiment of the present invention, the data set management module 100 is further specifically configured to: view the data set, and/or modify the name of the data set, and/or supplement the data.
That is, the data set management module 100 may perform operations such as viewing, modifying names, supplementing data, etc., on the uploaded data set in addition to creating a new data set.
It is noted that, based on the automated target detection platform, the underlying architecture diagram of the automated target detection platform according to an embodiment of the present invention can be shown in fig. 5. The underlying architecture of the automatic target detection platform may include an AI engine layer, an AI middleware and an AI framework layer, wherein a data center (data set management module 100) in the AI middleware corresponds to data input, data management and feature construction in the AI engine layer, a model center (model training module 200, model verification module 300, model publishing module 400 and model creating and querying module) in the AI middleware corresponds to automatic modeling, model iteration and model management in the AI engine layer, and an application center in the AI middleware corresponds to model application, application management and application monitoring in the AI engine layer, wherein data in application monitoring may also automatically flow back to data input in the AI engine layer. Correspondingly, the AI framework layer may include: the system comprises a high-dimensional computing framework, a container cloud, an application management tool, distributed computing, high-performance object storage, a data stream management module, distributed storage, an intelligent scheduling engine and automatic resource configuration. That is, the automatic target detection platform can implement the functions of the above embodiments based on the underlying architecture, and in particular, the above embodiments can be referred to, and details thereof are not described herein to avoid redundancy.
Therefore, the automatic target detection platform can strengthen basic support for artificial intelligence research and development application, especially forms an ecological chain for promoting mutual cooperation among artificial intelligence software, hardware and an intelligent cloud aiming at a target detection problem, and forms an intelligence and public creation platform and a service environment for innovative links for obstetrical and scientific research. Particularly, the practical problems to be faced by small and medium artificial intelligence entrepreneurship enterprises are solved, the technical threshold of the target detection algorithm is reduced, and the target detection algorithm based on deep learning is easier to use.
In summary, according to the automated target detection platform in the embodiment of the present invention, a data set is created and managed by the data set management module, a model to be trained is trained by the model training module according to the data set to generate a first target detection model, the first target detection model is verified on line by the model verification module to generate a second target detection model, and the second target detection model is released by the model release module. Therefore, the technical threshold of the deep learning target detection algorithm can be greatly reduced, and the door entry difficulty is reduced.
In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. The meaning of "plurality" is two or more unless specifically limited otherwise.
In the present invention, unless otherwise expressly stated or limited, the terms "mounted," "connected," "secured," and the like are to be construed broadly and can, for example, be fixedly connected, detachably connected, or integrally formed; can be mechanically or electrically connected; either directly or indirectly through intervening media, either internally or in any other relationship. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
In the present invention, unless otherwise expressly stated or limited, the first feature "on" or "under" the second feature may be directly contacting the first and second features or indirectly contacting the first and second features through an intermediate. Also, a first feature "on," "over," and "above" a second feature may be directly or diagonally above the second feature, or may simply indicate that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature may be directly under or obliquely under the first feature, or may simply mean that the first feature is at a lesser elevation than the second feature.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (7)
1. An automated target detection platform, comprising:
a dataset management module to create and manage datasets;
the model training module is used for training a model to be trained according to the data set so as to generate a first target detection model;
the model checking module is used for carrying out online checking on the first target detection model so as to generate a second target detection model;
and the model issuing module is used for issuing the second target detection model.
2. The automated target detection platform of claim 1, wherein the dataset management module is specifically configured to: labeling the target image to obtain a sample image, performing format conversion on the sample image, and compressing the sample image after format conversion to generate the data set.
3. The automated target detection platform of claim 2, wherein the data set management module is further specifically configured to:
view the data set, and/or
Modifying the name of the data set, and/or
And supplementing the data.
4. The automated target detection platform of claim 1, wherein the model publishing module is specifically configured to:
and issuing the second target detection model by adopting cloud service and/or offline service.
5. The automated target detection platform of claim 1, further comprising:
the identity authentication module is used for receiving identity information input by a user, authenticating the identity of the user according to the identity information, and opening the use permission of the user after the identity of the user is successfully authenticated.
6. The automated target detection platform of claim 1, further comprising:
and the model creating and inquiring module is used for creating the model to be trained and inquiring the model information of the second target detection model.
7. The automated target detection platform of claim 6, wherein the model information comprises a model version number, a model type, a release status, and a model effect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010924172.3A CN112070149A (en) | 2020-09-04 | 2020-09-04 | Automatic target detection platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010924172.3A CN112070149A (en) | 2020-09-04 | 2020-09-04 | Automatic target detection platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112070149A true CN112070149A (en) | 2020-12-11 |
Family
ID=73666539
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010924172.3A Pending CN112070149A (en) | 2020-09-04 | 2020-09-04 | Automatic target detection platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112070149A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508193A (en) * | 2021-02-02 | 2021-03-16 | 江西科技学院 | Deep learning platform |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109447248A (en) * | 2018-11-09 | 2019-03-08 | 深圳增强现实技术有限公司 | Deep learning platform and method |
| CN109508485A (en) * | 2018-10-30 | 2019-03-22 | 平安医疗健康管理股份有限公司 | A kind of data processing model dissemination method, device, server and storage medium |
| CN111310934A (en) * | 2020-02-14 | 2020-06-19 | 北京百度网讯科技有限公司 | Model generation method and device, electronic equipment and storage medium |
-
2020
- 2020-09-04 CN CN202010924172.3A patent/CN112070149A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109508485A (en) * | 2018-10-30 | 2019-03-22 | 平安医疗健康管理股份有限公司 | A kind of data processing model dissemination method, device, server and storage medium |
| CN109447248A (en) * | 2018-11-09 | 2019-03-08 | 深圳增强现实技术有限公司 | Deep learning platform and method |
| CN111310934A (en) * | 2020-02-14 | 2020-06-19 | 北京百度网讯科技有限公司 | Model generation method and device, electronic equipment and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508193A (en) * | 2021-02-02 | 2021-03-16 | 江西科技学院 | Deep learning platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR20190139938A (en) | Picture-based vehicle damage estimation method and apparatus, and electronic device | |
| CN111523413B (en) | Method and device for generating face image | |
| JP2022521038A (en) | Face recognition methods, neural network training methods, devices and electronic devices | |
| CN105051754A (en) | Method and apparatus for detecting people by a surveillance system | |
| CN113822247B (en) | Method and system for identifying illegal building based on aerial image | |
| CN111931047B (en) | Artificial intelligence-based black product account detection method and related device | |
| CN115049057B (en) | Model deployment method and device, electronic equipment and storage medium | |
| US20210192215A1 (en) | Platform for the management and validation of contents of video images, picture or similar, generated by different devices | |
| CN113469933A (en) | Generative antagonistic network model for small road object detection | |
| CN112233428B (en) | Traffic flow prediction method, device, storage medium and equipment | |
| CN107870673A (en) | Specification implementation method based on AR technologies | |
| CN114862372B (en) | Intelligent education data tamper-proof processing method and system based on block chain | |
| CN112333165A (en) | Identity authentication method, device, equipment and system | |
| CN112070149A (en) | Automatic target detection platform | |
| CN113516102A (en) | Deep learning parabolic behavior detection method based on video | |
| CN115131634A (en) | Image recognition method, device, equipment, storage medium and computer program product | |
| CN115830399A (en) | Classification model training method, apparatus, device, storage medium, and program product | |
| CN113158771B (en) | Community personnel information acquisition method, system and application | |
| CN112308093A (en) | Air quality perception method based on image recognition, model training method and system | |
| CN113593256B (en) | Unmanned aerial vehicle intelligent driving-away control method and system based on city management and cloud platform | |
| CN110414792A (en) | Component centralized procurement management system and Related product based on BIM and big data | |
| CN112769929B (en) | Site-to-site loop test system and method for vehicle-road cooperation technology | |
| CN111062479B (en) | Neural network-based rapid model upgrading method and device | |
| CN113656842A (en) | Data verification method, device and equipment | |
| CN113011223A (en) | Image recognition method, system, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |