A data set storage method for face recognition test system
A face recognition and testing system technology, applied in database indexing, structured data retrieval, special data processing applications, etc., can solve problems that cannot meet large-scale storage applications, and achieve the effect of improving operating performance
Active Publication Date: 2022-07-26
ZHEJIANG UNIV OF TECH
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Traditional network storage systems use centralized storage servers to store all data. Storage servers become the bottleneck of system performance and the focus of reliability and security, which cannot meet the needs of large-scale storage applications.
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
Embodiment Construction
[0029] The present invention will be further described below in conjunction with the accompanying drawings:
[0030] refer to Figure 1 to Figure 3 , a method for storing data sets of a face recognition test system, comprising the following steps:
[0031] Step 1. Customized datasets: Crawling multiple public face datasets (CAS-PEAL, WIDER FACE, LabeledFaces in the Wild, YouTube Faces DB, CASIA-SURF), according to the three scenarios of face recognition SDK application (face 1:1 recognition function, face 1:N recognition function, special face recognition function), the data set is divided into fourteen attributes, namely live detection, race, time span, 3D angle detection, illumination angle, multi-adornment, Background change, small face detection, multiple face detection, age recognition, blur detection, occlusion detection, gender recognition and video verification;
[0032] Step 2. Data set index unification: Design a 64-bit numbering structure for image naming, design ...
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
A data set storage method for a face recognition test system, comprising the following steps: Step 1: Customizing a data set: crawling multiple public face data sets, and dividing the data set into fourteen sets according to three scenarios of face recognition SDK application attributes; step 2, data set indicators are unified; step 3, the file name is maintained as metadata in redis-cluster: the data set is stored in different individual storage partitions, and all data set information is maintained in the redis-cluster cluster, in the redis-cluster Use the Hash structure to allocate the data sets collected online by name, and store the metadata information such as the data set storage path and image file name in the Set collection; Step 4 uses the Redis Hash data structure to store key-value data. The purpose of the present invention is to improve the running performance of the test system, and to achieve the final effect of improving the face recognition system according to the deficiencies.
Description
technical field [0001] The invention belongs to the field of computer big data, and designs a 10W level data set that can be used for testing a face recognition system and a storage method thereof based on distributed storage technology. Background technique [0002] With the continuous advancement of technologies in the fields of artificial intelligence and machine learning, and the promotion of the national "new infrastructure" strategy, new technologies and applications have accelerated the digital transformation of traditional industries, and data has grown exponentially. Massive data creates infinite value in being analyzed and mined. [0003] Internet companies have invested heavily in data storage and management, and increased the construction of data centers. Ali invested 200 billion in the research and development of cloud operating systems, servers, chips and other technologies to cope with the arrival of massive data. Tencent took advantage of the victory to inves...
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 Patents(China)
IPC IPC(8): G06F16/22
CPCG06F16/2255
Inventor 郑薇胡海根邱天章远
Owner ZHEJIANG UNIV OF TECH