CN111680736A - Artificial intelligence behavior analysis model training system and method based on virtual reality - Google Patents
Artificial intelligence behavior analysis model training system and method based on virtual reality Download PDFInfo
- Publication number
- CN111680736A CN111680736A CN202010492931.3A CN202010492931A CN111680736A CN 111680736 A CN111680736 A CN 111680736A CN 202010492931 A CN202010492931 A CN 202010492931A CN 111680736 A CN111680736 A CN 111680736A
- Authority
- CN
- China
- Prior art keywords
- virtual
- analysis model
- data
- environment
- behavior analysis
- 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
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/40—Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
-
- 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
- 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
Abstract
The invention discloses an artificial intelligence behavior analysis model training system and method based on virtual reality, which utilize the virtual reality technology to construct various simulation scene environment simulation realization scenes, and rapidly generate a large amount of simulation data which is real enough and has marks by controlling a three-dimensional human body, vehicle or other object model with mark points to carry out a series of behavior actions. And a convenient and simple training test environment is provided for behavior analysis model optimization and model deployment. The problems of difficulty in data collection, low data quality, high data marking cost, low training efficiency and the like in the training process of the behavior analysis model are solved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence behavior analysis model training system and method based on virtual reality.
Background
With the rapid development of informatization, artificial intelligence technologies based on vision, such as face recognition, image retrieval, target detection and tracking, and the like, have covered various occasions. Behavior analysis, as a higher-level artificial intelligence technology, integrates target detection and tracking, target (mainly human) motion capture, and behavior judgment integrating environmental information, and is applied to a wider field. Because modern artificial intelligence is mainly driven by data, obtaining high-quality data is a serious point for training a behavior analysis algorithm to be algorithm research, algorithm optimization and application landing. But the key problems of difficult data acquisition, poor data quality and the like exist in the application of specific occasions.
The virtual reality technology can be just used as a simulation data generator to be applied to behavior analysis model training and system design, along with the deep research, the virtual reality technology is widely applied to a simulation environment and has wide application in the fields of automatic driving, industry and medical treatment.
Chinese patent document CN100589148C discloses a training-oriented automobile driving simulator implementation method. The virtual visual, sound effect and motion simulation in the automobile driving process are generated by a computer by means of computer graphics, virtual reality and other technologies, so that a driving learner is immersed in a virtual driving environment, a college thinks of driving actions of the virtual driving environment according to visual, auditory and tactile feelings provided by the virtual driving environment, and an operating mechanism in a simulated cockpit is operated. The invention adapts to the increasing demand of driving training, saves resources such as fields, automobiles, gasoline, manpower and the like, has various examination items and error prompt functions, enables users to master basic driving skills in a safer environment, can achieve the effects of familiarizing traffic rules and perfecting driving skills when practicing in urban scenes, mountain roads and expressways, can simulate and train the driving skills of the users under severe conditions and the strain capacity to various emergency conditions in various weathers, and has important application value in the fields of commerce and scientific research.
Chinese patent document CN101667346B discloses an amputation upper limb rehabilitation training system based on virtual reality, which includes myoelectric signal detection and processing, amputation upper limb modeling, and virtual reality scene interaction. The myoelectric signal detection and processing part utilizes a myoelectric tester to extract, amplify, filter, convert A/D and collect multiple paths of myoelectric signals on the surface of the stump, extracts the robust characteristics of the myoelectric signals and combines an online learning method to realize quick and effective upper limb movement action recognition; the amputation upper limb modeling part carries out three-dimensional reconstruction on the amputation upper limb by using a picture of a healthy upper limb and a three-dimensional parameterized grid model, and realizes the action simulation of a virtual hand by using the upper limb motion parameters obtained by tracking as model driving data; the virtual reality scene interaction part carries out real three-dimensional interaction scene modeling, and realizes real-time interaction between the residual limb muscle action and the three-dimensional scene through myoelectricity. The device is mainly used for assisting an upper limb amputation patient to perform necessary adaptive training before the artificial limb is installed, so as to help the patient to adapt to the use of the artificial limb as soon as possible.
However, the main service object of the virtual reality technology is a person, and is not an intelligent body, and most of the existing virtual reality technology is applied to various scene simulations, but does not concentrate on artificial intelligence training by using simulation data, and in the training process of an artificial intelligence model, especially a behavior recognition model, a large amount of data needs to be provided for specific actions or behaviors for training, if a camera is used for collecting real scenes, the data acquisition process is complex, the data quality is uneven, the cost is high, and the efficiency is low.
The images of the behavior actions generated by the virtual reality technology can provide data aiming at specific actions, and the training requirements of the behavior recognition model are fully met, but the artificial intelligence model training is not carried out in a virtual environment by the current technology.
Disclosure of Invention
The invention aims to design and develop an artificial intelligence behavior analysis model training system based on virtual reality, construct various simulation scene environment simulation realization scenes by utilizing the virtual reality technology, rapidly generate a large amount of marked simulation data by controlling a three-dimensional human body, vehicle or other object model with mark points to perform a series of behavior actions, and provide a convenient and simple training test environment for behavior analysis model optimization and model deployment.
The invention also aims to design and develop an artificial intelligence behavior analysis model training method based on virtual reality, which can randomly switch visual angles in a virtual environment to obtain various virtual data and train a behavior analysis network model, so that the model can be quickly verified, and environments which are difficult to simulate under more real scenes can be constructed.
The technical scheme provided by the invention is as follows:
a virtual reality-based artificial intelligence behavioral analysis model training system, comprising:
the client is in a B/S style and is provided with a data interface for building a virtual environment, inserting a virtual target, acquiring virtual data and building a behavior analysis model;
and the central server is interacted with the client and is used for behavior analysis model training, behavior analysis model testing, simulation environment rendering and training result rendering.
Preferably, the client includes:
the virtual environment building module is used for building a simulation environment;
the virtual target generation module is connected with the virtual environment building module and is used for setting virtual target information and inserting a virtual target into the simulation environment;
and the visual operation interface is connected with the virtual environment building module and the virtual target generating module and is used for receiving the simulation environment state and the virtual target information.
Preferably, the central server includes:
the data generation module is connected with the client and used for providing training virtual data for the behavior analysis model;
and the label generation module is connected with the data generation module and is used for acquiring virtual data.
Preferably, the data generation module includes:
a training set which trains a behavior analysis model in a virtual environment to generate trained virtual data;
and the test set monitors the effect of the behavior analysis model in the simulation environment and generates tested virtual data.
Preferably, the method further comprises the following steps:
and the virtual reality engine is arranged in the virtual environment building module, the virtual target generating module and the data generating module and can provide simulation environments with different simulation degrees and different environment information.
Preferably, the virtual target information includes: number of virtual targets, motion, and travel speed.
Preferably, the virtual environment building module prestores a common simulation environment, and the virtual target generating module prestores a virtual target.
Preferably, the virtual data includes: the method comprises the steps of 3D point cloud, video and corresponding sequence images, camera parameters, 2D detection frames, semantic segmentation, instance segmentation, panoramic segmentation labels, 2D character key points, 3D detection frames, 3D labels, 3D character key points, video labels corresponding to targets, action occurrence space areas and action occurrence time areas.
A training method of artificial intelligence behavior analysis model based on virtual reality uses the training system of artificial intelligence behavior analysis model based on virtual reality, including the following steps:
step one, building a virtual environment;
step two, setting a virtual target, and inserting the virtual target into the virtual environment;
selecting an acquisition visual angle;
and step four, performing behavior analysis model training and testing on the virtual target to obtain virtual data.
The invention has the following beneficial effects:
the artificial intelligence behavior analysis model training system based on the virtual reality can realize a customized virtual simulation environment through the virtual reality technology, acquire virtual data in the environment, provide a large number of basic artificial intelligence task models for developers, greatly improve the problems that the conventional artificial intelligence behavior analysis task is difficult to acquire data and has high cost, accelerate the research and development efficiency, and reduce the research and development cost.
The artificial intelligence behavior analysis model training system based on the virtual reality is composed of a central server and a client, is simple in structure, easy to deploy, low in cost and easy to upgrade and maintain, generally only needs to upgrade software of the central server and the client, is low in upgrade and maintenance cost, can reduce the difficulty, time and hardware cost of acquiring complex scene data, and has strong use value and ideal application prospect.
The artificial intelligence behavior analysis model training method based on virtual reality can switch the visual angle at will in the virtual environment to obtain various virtual data and train the behavior analysis network model, so that the model can be verified quickly, and environments which are difficult to simulate under more real scenes can be constructed.
Drawings
Fig. 1 is a schematic block diagram of a training system for an artificial intelligence behavior analysis model based on virtual reality according to the present invention.
Detailed Description
The present invention is described in further detail below in order to enable those skilled in the art to practice the invention with reference to the description.
As shown in fig. 1, the artificial intelligence behavior analysis model training system based on virtual reality provided by the present invention is mainly composed of two layers, namely, a client and a central server, wherein the client includes: the system comprises a virtual environment building module, a virtual target generating module and a visual operation interface; the center server includes: the system comprises a data (virtual training data and test data) generation module, a label (virtual training label and test label) generation module and a storage module.
A virtual environment building module: the simulation environment with different simulation degrees is provided by building the simulation environment through the virtual reality engine, the simulation environment is used for simulating environment abstractions with different levels, and common simulation environments are prestored in the virtual environment building module.
The user can carry out 3D space modeling through the virtual reality engine and build a simulation environment by oneself, also can use the commonly used simulation environment of building in advance to carry out the environment and build to the user can switch information adjustment simulation environment's sense of reality such as light, weather of environment through 3D space template.
A virtual target generation module: the method is realized by inserting a virtual target in a simulation environment, and the virtual target is prestored in the virtual target generation module.
The virtual target includes: pedestrian, vehicle and unmanned aerial vehicle, the user can self-define virtual target, also can use the prefabricated virtual target of system, and the user can define information such as quantity, action and the speed of travel of virtual target simultaneously.
Virtual environment construction and target insertion are realized through virtual reality engine rendering, virtual images or videos are generated, and the virtual target is inserted through embedding a target file into a constructed 3D space.
And the virtual reality engine is arranged in the virtual environment building module, the virtual target generating module and the data generating module and can provide simulation environments with different simulation degrees and different environment information.
And the plurality of acquisition devices are arranged in the virtual environment building module and used for acquiring virtual reality simulation under different viewing angles, and the plurality of acquisition devices are also connected with the data generation module and used for outputting acquired data.
Preferably, the acquisition device is a camera. Through setting up different camera positions, obtain the virtual reality emulation under the different visual angles to can export and include: 3D point clouds, video and corresponding sequence images.
A data generation module for generating virtual data and providing training data for a behavior analysis model provided or selected by a user, the data generation module comprising:
a training set which trains a behavior analysis model in a virtual environment to generate trained virtual data;
and the test set monitors the effect of the behavior analysis model in the simulation environment and generates tested virtual data.
A user needs to set a virtual environment and a target through a virtual environment building module and a virtual target generating module, and then virtual training data and test data are generated through a data generating module.
A tag generation module: virtual data at the corresponding view angle is generated.
The behavior analysis model is provided by a user, and the behavior analysis model for the experiment generates a training set and a test set by using a virtual environment and a target. And training the behavior analysis model by using the generated training set, and detecting the effect of the model by using the test set after training is finished.
The virtual data includes: camera parameters; 2D detection boxes, semantic segmentation, instance segmentation, panoramic segmentation labels and 2D character key points; the system comprises a 3D detection frame, a 3D label and 3D character key points; and corresponding video tags, spatial regions of motion occurrence, temporal regions of motion occurrence, and the like.
The visual operation interface is a B/S style client based on a Web interface and is mainly used for a PC (personal computer) terminal; and a visual interface, a programmable interface and a general data input and output interface are provided, so that the user can conveniently manage and modify the simulation environment.
And the storage module is used for storing all virtual data, namely the virtual data is a hard disk.
The invention relates to a virtual reality-based artificial intelligence behavior analysis model training method, which comprises the following steps:
the user needs to set a virtual environment and a virtual target at the client, and embed a user-defined target file into the virtual environment, after the virtual environment and the target are established, the user can set a camera view angle at any position in the virtual environment, render a picture according to the camera position, generate an image or a video, and select data to be generated, such as: camera parameters at the corresponding view angle; 2D detection boxes, semantic segmentation, instance segmentation, panoramic segmentation labels and 2D character key points; the system comprises a 3D detection frame, a 3D label and 3D character key points; and the corresponding video label, the action occurrence space area and the action occurrence time area. These data will be stored in json format in a storage module in the central server. And deploying a deep learning network framework on the cloud server, wherein a behavior analysis model provided by a user is the same as the deployed deep learning network framework, then performing deep learning training by using the generated training set and the label, and simultaneously rendering the virtual environment and returning the rendered virtual environment to the client by the central server. A user can design an artificial intelligence model for behavior analysis and edit in the system, and the artificial intelligence model for behavior analysis is trained and verified by utilizing virtual simulation data.
And the central server side renders the virtual environment and stores the virtual data, and simultaneously acquires the data label required by the behavior analysis model. The user will deploy the behavioral analysis model to the central server. The central server will utilize computing resources to implement behavioral analysis model training. And finally, rendering the model test result to virtual data, so that visual understanding is facilitated.
The invention relates to an artificial intelligence behavior analysis model training system based on virtual reality, which is designed and developed by the invention, utilizes the virtual reality technology to construct various simulation scene environment simulation realization scenes, and rapidly generates a large amount of simulation data which is real enough and has marks by controlling a three-dimensional human body, vehicle or other object model with the marks to perform a series of behavior actions. And a convenient and simple training test environment is provided for behavior analysis model optimization and model deployment. The problems of difficulty in data collection, low data quality, high data marking cost, low training efficiency and the like in the training process of the behavior analysis model are solved.
The artificial intelligence behavior analysis model training method based on virtual reality can switch the visual angle at will in the virtual environment to obtain various virtual data and train the behavior analysis network model, so that the model can be verified quickly, and environments which are difficult to simulate under more real scenes can be constructed.
While embodiments of the invention have been described above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable to various fields of endeavor for which the invention may be embodied with additional modifications as would be readily apparent to those skilled in the art, and the invention is therefore not limited to the details given herein and to the embodiments shown and described without departing from the generic concept as defined by the claims and their equivalents.
Claims (9)
1. An artificial intelligence behavioral analysis model training system based on virtual reality, comprising:
the client is in a B/S style and is provided with a data interface for building a virtual environment, inserting a virtual target, acquiring virtual data and building a behavior analysis model;
and the central server is interacted with the client and is used for behavior analysis model training, behavior analysis model testing, simulation environment rendering and training result rendering.
2. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 1, wherein the client comprises:
the virtual environment building module is used for building a simulation environment;
the virtual target generation module is connected with the virtual environment building module and is used for setting virtual target information and inserting a virtual target into the simulation environment;
and the visual operation interface is connected with the virtual environment building module and the virtual target generating module and is used for receiving the simulation environment state and the virtual target information.
3. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 2, wherein the central server includes:
the data generation module is connected with the client and used for providing training virtual data for the behavior analysis model;
and the label generation module is connected with the data generation module and is used for acquiring virtual data.
4. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 3, wherein the data generation module includes:
a training set which trains a behavior analysis model in a virtual environment to generate trained virtual data;
and the test set monitors the effect of the behavior analysis model in the simulation environment and generates tested virtual data.
5. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 4, further comprising:
and the virtual reality engine is arranged in the virtual environment building module, the virtual target generating module and the data generating module and can provide simulation environments with different simulation degrees and different environment information.
6. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 5, wherein the virtual goal information includes: number of virtual targets, motion, and travel speed.
7. The virtual reality-based artificial intelligence behavior analysis model training system of claim 6, wherein the virtual environment building module is pre-stored with a common simulation environment, and the virtual target generation module is pre-stored with a virtual target.
8. The virtual reality-based artificial intelligence behavioral analysis model training system according to claim 7, wherein the virtual data includes: the method comprises the steps of 3D point cloud, video and corresponding sequence images, camera parameters, 2D detection frames, semantic segmentation, instance segmentation, panoramic segmentation labels, 2D character key points, 3D detection frames, 3D labels, 3D character key points, video labels corresponding to targets, action occurrence space areas and action occurrence time areas.
9. A virtual reality-based artificial intelligence behavior analysis model training method using the virtual reality-based artificial intelligence behavior analysis model training system according to any one of claims 1 to 8, comprising the steps of:
step one, building a virtual environment;
step two, setting a virtual target, and inserting the virtual target into the virtual environment;
selecting an acquisition visual angle;
and step four, performing behavior analysis model training and testing on the virtual target to obtain virtual data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010492931.3A CN111680736A (en) | 2020-06-03 | 2020-06-03 | Artificial intelligence behavior analysis model training system and method based on virtual reality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010492931.3A CN111680736A (en) | 2020-06-03 | 2020-06-03 | Artificial intelligence behavior analysis model training system and method based on virtual reality |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111680736A true CN111680736A (en) | 2020-09-18 |
Family
ID=72434981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010492931.3A Pending CN111680736A (en) | 2020-06-03 | 2020-06-03 | Artificial intelligence behavior analysis model training system and method based on virtual reality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111680736A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990136A (en) * | 2021-04-29 | 2021-06-18 | 成都深蓝思维信息技术有限公司 | Target detection method and device |
| CN113673172A (en) * | 2021-09-06 | 2021-11-19 | 中国农业大学 | Corn ear high-flux intelligent screening virtual simulation experiment system |
| CN114019921A (en) * | 2021-11-01 | 2022-02-08 | 碳纪科技(北京)有限公司 | Production line equipment control method, device, equipment and medium based on combined model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130125028A1 (en) * | 2011-11-11 | 2013-05-16 | Cobham CTS Ltd. | Hazardous Device Detection Training System |
| CN110837697A (en) * | 2019-10-25 | 2020-02-25 | 华南理工大学 | Intelligent traffic simulation system and method for intelligent vehicle |
| CN111047708A (en) * | 2019-11-26 | 2020-04-21 | 中国人民解放军空军工程大学 | Complex equipment high-risk project training system based on mixed reality |
-
2020
- 2020-06-03 CN CN202010492931.3A patent/CN111680736A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130125028A1 (en) * | 2011-11-11 | 2013-05-16 | Cobham CTS Ltd. | Hazardous Device Detection Training System |
| CN110837697A (en) * | 2019-10-25 | 2020-02-25 | 华南理工大学 | Intelligent traffic simulation system and method for intelligent vehicle |
| CN111047708A (en) * | 2019-11-26 | 2020-04-21 | 中国人民解放军空军工程大学 | Complex equipment high-risk project training system based on mixed reality |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990136A (en) * | 2021-04-29 | 2021-06-18 | 成都深蓝思维信息技术有限公司 | Target detection method and device |
| CN112990136B (en) * | 2021-04-29 | 2021-08-03 | 成都深蓝思维信息技术有限公司 | Target detection method and device |
| CN113673172A (en) * | 2021-09-06 | 2021-11-19 | 中国农业大学 | Corn ear high-flux intelligent screening virtual simulation experiment system |
| CN114019921A (en) * | 2021-11-01 | 2022-02-08 | 碳纪科技(北京)有限公司 | Production line equipment control method, device, equipment and medium based on combined model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111680736A (en) | Artificial intelligence behavior analysis model training system and method based on virtual reality | |
| Shan et al. | Research on landscape design system based on 3D virtual reality and image processing technology | |
| CN102053563A (en) | Flight training data acquisition and quality evaluation system of analog machine | |
| CN107423398A (en) | Exchange method, device, storage medium and computer equipment | |
| CN106601060A (en) | Virtual reality system for experiencing fire-fighting scene | |
| US20050275717A1 (en) | Method and apparatus for testing stereo vision methods using stereo imagery data | |
| CN109783914B (en) | Preprocessing dynamic modeling method and device based on virtual reality simulation | |
| CN110827602A (en) | Cable joint manufacturing and operation and maintenance skill training device and method based on VR + AR technology | |
| Ganoni et al. | A framework for visually realistic multi-robot simulation in natural environment | |
| Yu et al. | Intelligent visual-IoT-enabled real-time 3D visualization for autonomous crowd management | |
| Wu | RETRACTED: Design of embedded dance teaching control system based on FPGA and motion recognition processing | |
| Zhang et al. | Nerf-lidar: Generating realistic lidar point clouds with neural radiance fields | |
| CN115186473A (en) | Scene perception modeling and verifying method based on parallel intelligence | |
| Xu et al. | Reliability of gan generated data to train and validate perception systems for autonomous vehicles | |
| CN112686979B (en) | Simulated pedestrian animation generation method and device and electronic equipment | |
| CN104933278B (en) | A kind of multi-modal interaction method and system for disfluency rehabilitation training | |
| CN116957866A (en) | Individualized teaching device of digital man teacher | |
| Ballout et al. | The benefits of synthetic data for action categorization | |
| CN115202479A (en) | Man-machine co-driving simulation system of man-in-loop and application thereof | |
| Jiang | RETRACTED: Research on auxiliary methods of swimming training virtual simulation technology based on embedded computer✰ | |
| Aranjuelo et al. | Leveraging Synthetic Data for DNN-Based Visual Analysis of Passenger Seats | |
| CN112883785A (en) | Railway operation environment safety image detection system based on parallel vision theory | |
| Azam et al. | An HMD-based pedestrian simulator for training and measuring individuals perceptual-motor behaviour in road crossing | |
| Li | Convolutional Neural Network-Based Virtual Reality Real-Time Interactive System Design for Unity3D | |
| Chao et al. | Analysis of the Current Situation of Teaching and Learning of Ideological and Political Theory Courses by Deep Learning |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200918 |
|
| RJ01 | Rejection of invention patent application after publication |