Artificial intelligence computer vision reasoning method
A technology of computer vision and reasoning methods, applied in the field of artificial intelligence, can solve problems such as model call crashes affecting efficiency, achieve rapid deployment and application, easy maintenance, and improve the efficiency of computing resources
Pending Publication Date: 2022-04-08
广州中科智巡科技有限公司
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, in practical applications, problems such as resource scheduling conflicts and model call crashes have greatly affected efficiency. In order to use and schedule models and resources more stably, securely and conveniently
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
no. 1 example
[0052] see figure 1 , a first embodiment of an artificial intelligence computer vision reasoning method provided by the present invention includes:
[0053] 101. Configure the algorithm engine.
[0054] In this embodiment, an algorithm engine that provides runtime support for algorithm reasoning is configured in the server cluster. The algorithm engine is developed based on C#.net core. The system uses k8s for resource scheduling management to generate corresponding program containers, and each user instance and scheduling unit run independently, avoiding program exceptions caused by mutual direct environment configuration.
[0055] According to the number of GPUs requested by the user, use the k8s command line tool to create a corresponding running container agent and listen to port 5000. The agent only supports the service startup method. Regardless of the input parameters, the agent will start a Web Server and listen to port 5000 (it can be mapped to other ports through...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an artificial intelligence computer visual reasoning method, which uses a k8s (k8s for short) cluster to manage a plurality of containers for running sub-tasks, has the advantage of easy maintenance, can quickly train operation and maintenance personnel, and achieves the purposes of improving the overall underlying computing resource efficiency and quickly deploying, operating and maintaining. The artificial intelligence computer vision reasoning method comprises the following steps: configuring an algorithm engine, wherein the algorithm engine is developed based on C #. Net core; the system uses k8s to perform resource scheduling management to generate a corresponding program container, and uses a k8s command line tool to create a corresponding running container agent according to the number of GPUs applied by a user; the running container agent monitors algorithm reasoning task information sent by a Web Api request received by a 5000 port; and the algorithm engine creates a script according to the algorithm reasoning task and executes an algorithm process.
Description
technical field [0001] The invention relates to the technical fields of artificial intelligence such as computer vision and deep learning, and in particular to an artificial intelligence computer vision reasoning method. Background technique [0002] The existing GPU computing resource scheduling scheme, one is to use the GPU resource scheduler for shared memory, combined with the cluster job management system, to pass the number of the required GPU device to the user, which can automatically allocate the required GPU for the user job Device resources to avoid usage conflicts of multiple GPU devices. Another GPU resource scheduling method and device for a MESOS containerized platform. When creating a target service on a graphic interface, by setting the GPU memory usage rate, it is convenient for the containerized platform to allocate the GPU memory usage rate for the target service. The corresponding target host, and by running the application container engine, calling the...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06N5/04G06T1/20
Inventor 范亮汤坚张磊
Owner 广州中科智巡科技有限公司