CN112464740A - Image processing method and system for top-down gesture recognition process - Google Patents
Image processing method and system for top-down gesture recognition process Download PDFInfo
- Publication number
- CN112464740A CN112464740A CN202011225263.4A CN202011225263A CN112464740A CN 112464740 A CN112464740 A CN 112464740A CN 202011225263 A CN202011225263 A CN 202011225263A CN 112464740 A CN112464740 A CN 112464740A
- Authority
- CN
- China
- Prior art keywords
- boundary frame
- image
- original
- gesture recognition
- image processing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000008569 process Effects 0.000 title claims abstract description 41
- 238000003672 processing method Methods 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 claims description 14
- 240000004282 Grewia occidentalis Species 0.000 claims description 5
- 238000012549 training Methods 0.000 abstract description 10
- 238000010586 diagram Methods 0.000 description 9
- 230000015654 memory Effects 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 3
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an image processing method and system for a top-down gesture recognition process, wherein the image processing method comprises the following steps: acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage; expanding the boundary frame in the original image to enable the expanded boundary frame to contain the whole character; adjusting the image with the expanded bounding box to a size with a preset size; and conveying the image after the size adjustment to a preset single posture recognition model for posture recognition. The image processing method for the top-down gesture recognition process is added in the traditional character gesture recognition process, so that the character recognition accuracy can be improved, the training difficulty of a subsequent gesture recognition model is reduced, and the generalization capability of the model to a scene is improved.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an image processing method and system for a top-down gesture recognition process.
Background
In the field of gesture recognition, the recognition process can be divided into single-person gesture recognition and multi-person gesture recognition according to the number of people in the recognized image. The single posture recognition is relatively simple in implementation process because different people in the image do not need to be distinguished, the current technical maturity is more mature compared with the multi-person posture recognition, and the multi-person posture recognition problem is the key point of research in the current posture recognition field and is also the direction of technology improvement in the future.
There are two mainstream research methods for multi-person gesture recognition, from top to bottom: detecting a plurality of persons, and then carrying out posture estimation on each person, wherein the posture estimation can be realized by adding single posture estimation on the basis of common person identification; from bottom to top: the joint points are detected firstly, and then the person to which each joint point belongs is judged.
In the top-down gesture recognition research process, for the first-stage character recognition, no matter using the traditional image recognition algorithm or the deep learning method, the positioning and recognition of the human body are highly likely to have the errors of incomplete character recognition, which is specifically represented as: the human body boundary in the human recognition result is outside the boundary box, for example, the palm of the human body with two open arms is most easily outside the boundary box, and the following reasons may be considered: firstly, partial image labeling results of the data set do not meet requirements when the data set is labeled, and all human body parts in a training set are not subjected to frame selection and labeling; secondly, the data set is huge, and the training times of the model are not enough; and thirdly, the model has partial defects and needs to be optimized. The errors inevitably increase the training difficulty of the second-stage single-person gesture recognition network and even cause final gesture recognition errors; in addition, the generalization capability of the whole network to the scene is also reduced due to the inconsistent distance between the person and the shot.
Disclosure of Invention
The invention provides an image processing method and system for a top-down gesture recognition process, which are used for solving the technical problems of incomplete character recognition and different character distances and lens distances under different scenes in the top-down gesture recognition process, so that the accuracy of the top-down gesture recognition process and the whole process is improved, the training difficulty of a subsequent gesture recognition model is reduced, and the generalization capability of the model to the scenes is improved.
In order to solve the technical problems, the invention provides the following technical scheme:
in one aspect, the present invention provides an image processing method for a top-down gesture recognition process, the image processing method for the top-down gesture recognition process comprising:
acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage;
expanding the boundary frame in the original image to enable the expanded boundary frame to contain the whole character;
adjusting the image with the expanded bounding box to a size with a preset size;
and conveying the image after the size adjustment to a preset single posture recognition model for posture recognition.
Further, expanding the bounding box in the original image includes:
calculating the coordinate of the center point O of the original boundary frame and the height H and the width W of the original boundary frame according to the four-corner coordinates of the original boundary frame in the original image;
with O as an origin, expanding the original boundary frame to (a.max (H, W))/2 towards two sides in the height direction and the width direction respectively to obtain the boundaries of four sides of the new boundary frame; taking the boundaries of the four sides of the new boundary frame as four parallel lines of the corresponding sides of the primary boundary frame to obtain an expanded boundary frame; where max (H, W) represents the maximum value of H and W, and a represents a preset coefficient.
Further, the value range of a is as follows: a is more than or equal to 1 and less than or equal to 1.5.
In another aspect, the present invention provides an image processing system for a top-down gesture recognition process, the image processing system for a top-down gesture recognition process comprising:
the input module is used for acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage;
the boundary frame expansion module is used for expanding the boundary frame in the original image, so that the expanded boundary frame comprises the whole person;
the size adjusting module is used for adjusting the image after the boundary frame is expanded to a size with a preset size;
and the output module is used for conveying the image after the size adjustment to a preset single posture recognition model so as to perform posture recognition.
Further, the bounding box extension module is specifically configured to:
calculating the coordinate of the center point O of the original boundary frame and the height H and the width W of the original boundary frame according to the four-corner coordinates of the original boundary frame in the original image;
with O as an origin, expanding the original boundary frame to (a.max (H, W))/2 towards two sides in the height direction and the width direction respectively to obtain the boundaries of four sides of the new boundary frame; taking the boundaries of the four sides of the new boundary frame as four parallel lines of the corresponding sides of the primary boundary frame to obtain an expanded boundary frame; where max (H, W) represents the maximum value of H and W, and a represents a preset coefficient.
Further, the value range of a is as follows: a is more than or equal to 1 and less than or equal to 1.5.
In yet another aspect, the present invention also provides an electronic device comprising a processor and a memory; wherein the memory has stored therein at least one instruction that is loaded and executed by the processor to implement the above-described method.
In yet another aspect, the present invention also provides a computer-readable storage medium having at least one instruction stored therein, the instruction being loaded and executed by a processor to implement the above method.
The technical scheme provided by the invention has the beneficial effects that at least:
firstly, adjusting the boundary frame on the original image according to the character boundary frame characteristics identified in the character identification stage to obtain an expanded new boundary frame, thereby overcoming the problem of incomplete character identification; then, according to the subsequent use of the character recognition, the proper proportion is adopted to carry out size adjustment on the character recognition result processed in the first stage, so that the problem of inconsistent distances between the character lenses in the image recognition process is solved; the image processing method provided by the invention is added in the traditional character gesture recognition process, so that the character recognition accuracy can be increased, the training difficulty of a subsequent gesture recognition model is reduced, and the generalization capability of the model to a scene is increased.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a schematic overall flow chart of an image processing method for a top-down gesture recognition process according to an embodiment of the present invention;
fig. 2 is a schematic diagram illustrating an implementation principle of a bounding box adjusting phase according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
First embodiment
The embodiment provides an image processing method for a top-down gesture recognition process, which can be implemented by an electronic device, which can be a terminal or a server. The method is applied to the middle stage of the top-down gesture recognition process, namely, the image processing method of the embodiment is added under the condition that each figure boundary frame in the image is recognized, and the final result of the image processing method of the embodiment is input into a subsequent single gesture recognition model for further gesture recognition. Specifically, the execution flow of the image processing method for the top-down gesture recognition process is shown in fig. 1, and comprises the following steps:
s101, acquiring an original image of each person boundary frame in the image which is identified in the person identification stage;
s102, expanding the boundary frame in the original image to enable the expanded boundary frame to contain the whole character;
further, as shown in fig. 2, expanding the bounding box in the original image includes the following steps:
firstly, calculating the coordinate of a central point O of an original boundary frame and the height H and the width W of the original boundary frame according to the four-corner coordinates of the original boundary frame in an original image, wherein H > W in the embodiment;
then, multiplying the maximum value max (H, W) of H and W by a preset coefficient a to obtain a.max (H, W); wherein, the value range of a is as follows: a is more than or equal to 1 and less than or equal to 1.5; in this embodiment, max (H, W) ═ H, a ═ 1.1;
finally, with O as an origin, expanding the original boundary frame to (a.max (H, W))/2 towards two sides in the height direction and the width direction respectively to obtain the boundaries of four sides of the new boundary frame; taking the boundaries of the four sides of the new boundary frame as four parallel lines of the corresponding sides of the primary boundary frame to obtain an expanded boundary frame;
it should be noted that, considering that the human body is easily out of the bounding box in the transverse direction (such as open arms, extending feet or legs) in general, and can be completely contained in the longitudinal height direction (the top of the head or the sole do not affect the subsequent recognition effect even if partially out of the bounding box) in general, and the longitudinal height of the human body bounding box is higher than the transverse width in general, the new bounding box adjusted by the bounding box is more likely to contain all the human body parts than before the treatment, so that the problem of incomplete human recognition can be overcome.
S103, adjusting the image with the expanded bounding box to a size with a preset size;
it should be noted that, although the problem of incomplete character recognition has been overcome in the previous stage, the sizes of recognized human bodies are different due to different distances between characters and shots, and the difficulty of subsequent posture recognition network training is still large. Therefore, after the previous stage, the size adjustment processing is required to be introduced, the size of the human body image obtained by adjusting the bounding box is uniformly adjusted to the same size (for example, 500 × 500) so as to overcome the problem of inconsistent distances between the shots of the characters, and then the processed human body image is input into the subsequent posture recognition or other networks, so that the purposes of reducing the difficulty of the subsequent network training recognition and enhancing the scene generalization capability of the whole network are achieved.
And S104, conveying the image after size adjustment to a single posture recognition model for posture recognition.
The method of the embodiment is suitable for a top-down gesture recognition process, namely a process of recognizing each character first and then recognizing the gesture of each character. The method of the embodiment is added between the two stages, so that the problems of high training difficulty and low generalization capability of the model scene caused by incomplete character recognition in the first stage and inconsistent distance between the original character shots in the second stage can be solved. According to the method, the problems of incomplete character recognition and inconsistent character lens distance are combined and solved, the final accuracy of the whole gesture recognition process can be improved while the accuracy of each stage is improved, the overall training difficulty of the model is reduced, and the generalization capability of the model scene is improved.
Second embodiment
The embodiment provides an image processing system for a top-down gesture recognition process, which comprises the following modules:
the input module is used for acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage;
the boundary frame expansion module is used for expanding the boundary frame in the original image, so that the expanded boundary frame comprises the whole person;
the size adjusting module is used for adjusting the image after the boundary frame is expanded to a size with a preset size;
and the output module is used for conveying the image after the size adjustment to a preset single posture recognition model so as to perform posture recognition.
The image processing system for the top-down gesture recognition process of the present embodiment corresponds to the image processing method for the top-down gesture recognition process of the first embodiment described above; the functions realized by each functional module in the image processing system for the top-down gesture recognition process correspond to the flow steps in the image processing method for the top-down gesture recognition process one by one; therefore, it is not described herein.
Third embodiment
The present embodiment provides an electronic device, which includes a processor and a memory; wherein the memory has stored therein at least one instruction that is loaded and executed by the processor to implement the method of the first embodiment.
The electronic device may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) and one or more memories, where at least one instruction is stored in the memory, and the instruction is loaded by the processor and executes the method.
Fourth embodiment
The present embodiment provides a computer-readable storage medium, in which at least one instruction is stored, and the instruction is loaded and executed by a processor to implement the method of the first embodiment. The computer readable storage medium may be, among others, ROM, random access memory, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like. The instructions stored therein may be loaded by a processor in the terminal and perform the above-described method.
Furthermore, it should be noted that the present invention may be provided as a method, apparatus or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media having computer-usable program code embodied in the medium.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It should also be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
Finally, it should be noted that while the above describes a preferred embodiment of the invention, it will be appreciated by those skilled in the art that, once the basic inventive concepts have been learned, numerous changes and modifications may be made without departing from the principles of the invention, which shall be deemed to be within the scope of the invention. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Claims (6)
1. An image processing method for a top-down gesture recognition process, comprising:
acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage;
expanding the boundary frame in the original image to enable the expanded boundary frame to contain the whole character;
adjusting the image with the expanded bounding box to a size with a preset size;
and conveying the image after the size adjustment to a preset single posture recognition model for posture recognition.
2. An image processing method for a top-down gesture recognition process according to claim 1, wherein expanding the bounding box in the original image comprises:
calculating the coordinate of the center point O of the original boundary frame and the height H and the width W of the original boundary frame according to the four-corner coordinates of the original boundary frame in the original image;
with O as an origin, expanding the original boundary frame to (a.max (H, W))/2 towards two sides in the height direction and the width direction respectively to obtain the boundaries of four sides of the new boundary frame; taking the boundaries of the four sides of the new boundary frame as four parallel lines of the corresponding sides of the primary boundary frame to obtain an expanded boundary frame; where max (H, W) represents the maximum value of H and W, and a represents a preset coefficient.
3. The image processing method for a top-down gesture recognition process of claim 2, wherein a ranges from: a is more than or equal to 1 and less than or equal to 1.5.
4. An image processing system for a top-down gesture recognition process, comprising:
the input module is used for acquiring an original image of each figure bounding box in the image which is identified in the figure identification stage;
the boundary frame expansion module is used for expanding the boundary frame in the original image, so that the expanded boundary frame comprises the whole person;
the size adjusting module is used for adjusting the image after the boundary frame is expanded to a size with a preset size;
and the output module is used for conveying the image after the size adjustment to a preset single posture recognition model so as to perform posture recognition.
5. The image processing system for a top-down gesture recognition process of claim 4, wherein the bounding box extension module is specifically configured to:
calculating the coordinate of the center point O of the original boundary frame and the height H and the width W of the original boundary frame according to the four-corner coordinates of the original boundary frame in the original image;
with O as an origin, expanding the original boundary frame to (a.max (H, W))/2 towards two sides in the height direction and the width direction respectively to obtain the boundaries of four sides of the new boundary frame; taking the boundaries of the four sides of the new boundary frame as four parallel lines of the corresponding sides of the primary boundary frame to obtain an expanded boundary frame; where max (H, W) represents the maximum value of H and W, and a represents a preset coefficient.
6. The image processing system for a top-down gesture recognition process of claim 5, wherein a ranges from: a is more than or equal to 1 and less than or equal to 1.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011225263.4A CN112464740A (en) | 2020-11-05 | 2020-11-05 | Image processing method and system for top-down gesture recognition process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011225263.4A CN112464740A (en) | 2020-11-05 | 2020-11-05 | Image processing method and system for top-down gesture recognition process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112464740A true CN112464740A (en) | 2021-03-09 |
Family
ID=74825051
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011225263.4A Pending CN112464740A (en) | 2020-11-05 | 2020-11-05 | Image processing method and system for top-down gesture recognition process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112464740A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222902A (en) * | 2021-04-16 | 2021-08-06 | 北京科技大学 | No-reference image quality evaluation method and system |
| CN114067328A (en) * | 2021-11-19 | 2022-02-18 | 中国建设银行股份有限公司 | Text recognition method and device and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108062526A (en) * | 2017-12-15 | 2018-05-22 | 厦门美图之家科技有限公司 | A kind of estimation method of human posture and mobile terminal |
| CN109657631A (en) * | 2018-12-25 | 2019-04-19 | 上海智臻智能网络科技股份有限公司 | Human posture recognition method and device |
| CN110674874A (en) * | 2019-09-24 | 2020-01-10 | 武汉理工大学 | Fine-grained image identification method based on target fine component detection |
| CN110751056A (en) * | 2019-09-27 | 2020-02-04 | 湖北工业大学 | Pedestrian motion prediction method based on improved top-down method multi-person posture detection |
| WO2020024185A1 (en) * | 2018-08-01 | 2020-02-06 | SZ DJI Technology Co., Ltd. | Techniques for motion-based automatic image capture |
| CN111241940A (en) * | 2019-12-31 | 2020-06-05 | 浙江大学 | Remote control method of robot and human body boundary frame determination method and system |
| CN111368678A (en) * | 2020-02-26 | 2020-07-03 | Oppo广东移动通信有限公司 | Image processing method and related device |
-
2020
- 2020-11-05 CN CN202011225263.4A patent/CN112464740A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108062526A (en) * | 2017-12-15 | 2018-05-22 | 厦门美图之家科技有限公司 | A kind of estimation method of human posture and mobile terminal |
| WO2020024185A1 (en) * | 2018-08-01 | 2020-02-06 | SZ DJI Technology Co., Ltd. | Techniques for motion-based automatic image capture |
| CN109657631A (en) * | 2018-12-25 | 2019-04-19 | 上海智臻智能网络科技股份有限公司 | Human posture recognition method and device |
| CN110674874A (en) * | 2019-09-24 | 2020-01-10 | 武汉理工大学 | Fine-grained image identification method based on target fine component detection |
| CN110751056A (en) * | 2019-09-27 | 2020-02-04 | 湖北工业大学 | Pedestrian motion prediction method based on improved top-down method multi-person posture detection |
| CN111241940A (en) * | 2019-12-31 | 2020-06-05 | 浙江大学 | Remote control method of robot and human body boundary frame determination method and system |
| CN111368678A (en) * | 2020-02-26 | 2020-07-03 | Oppo广东移动通信有限公司 | Image processing method and related device |
Non-Patent Citations (1)
| Title |
|---|
| KINGHT 金: "Adjust size and position of bounding boxes while keeping it somewhat centered", pages 1 - 3, Retrieved from the Internet <URL:stackoverflow.com/questions/47580887/adjust-size-and-position-of-bounding-boxes-while-keeping-it-somewhat-centered> * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222902A (en) * | 2021-04-16 | 2021-08-06 | 北京科技大学 | No-reference image quality evaluation method and system |
| CN113222902B (en) * | 2021-04-16 | 2024-02-02 | 北京科技大学 | No-reference image quality evaluation method and system |
| CN114067328A (en) * | 2021-11-19 | 2022-02-18 | 中国建设银行股份有限公司 | Text recognition method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111027493B (en) | Pedestrian detection method based on deep learning multi-network soft fusion | |
| CN106599836B (en) | Multi-face tracking method and tracking system | |
| CN109598234B (en) | Key point detection method and device | |
| CN109446889B (en) | Object tracking method and device based on twin matching network | |
| CN110069985B (en) | Image-based target point position detection method and device and electronic equipment | |
| WO2016138838A1 (en) | Method and device for recognizing lip-reading based on projection extreme learning machine | |
| CN107871099A (en) | Face detection method and apparatus | |
| CN111626184B (en) | Crowd density estimation method and system | |
| CN113095254B (en) | Method and system for positioning key points of human body part | |
| CN112464740A (en) | Image processing method and system for top-down gesture recognition process | |
| CN110969110A (en) | Face tracking method and system based on deep learning | |
| CN111275005B (en) | Drawn face image recognition method, computer-readable storage medium and related device | |
| CN111178363B (en) | Character recognition method, character recognition device, electronic equipment and readable storage medium | |
| CN114118303B (en) | Face key point detection method and device based on prior constraint | |
| CN113378864B (en) | Method, device and equipment for determining anchor frame parameters and readable storage medium | |
| CN114387656B (en) | Face changing method, device, equipment and storage medium based on artificial intelligence | |
| CN113762049B (en) | Content identification method, content identification device, storage medium and terminal equipment | |
| CN110826534A (en) | Face key point detection method and system based on local principal component analysis | |
| CN108229257A (en) | A kind of face recognition features' parallel training method based on deep learning and Spark | |
| CN117912058A (en) | Cattle face recognition method | |
| CN110580462B (en) | Natural scene text detection method and system based on non-local network | |
| CN113743308B (en) | Face recognition method, device, storage medium and system based on feature quality | |
| CN117611600B (en) | Image segmentation method, system, storage medium and device | |
| Li et al. | Research based on improved SSD target detection algorithm | |
| CN111626305B (en) | Target detection method, device and equipment |
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 |