Face recognition algorithm evaluation method based on quality dimension
A face recognition and algorithm technology, applied in the field of image processing, can solve the problems of easy misjudgment, little reference, and difficult to give advantages description.
Active Publication Date: 2019-10-25
易诚高科(大连)科技有限公司
View PDF2 Cites 12 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0007]1) For the same face recognition algorithm, the recognition rate is relatively the simplest quantitative index, and it cannot reflect in which situations the algorithm performs poorly and which situations The performance is good, and these situations are often the shortcomings of the algorithm design. In order to more objectively reflect the advantages and disadvantages of the algorithm, it is not comprehensive to measure the recognition rate only by multiple test libraries, and the reference value is not significant;
[0008]2) For different face recognition algorithms, the mainstream (state of the art) algorithm has little difference in recognition rate, and the recognition rate is used as a reference standard. It is not easy to give a relatively more detailed description of advantages, which makes the algorithm prone to misjudgment in the selection of application scenarios
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
[0069] The databases currently used for face testing have high diversity in terms of data volume, individual characteristics, postures, shooting equipment, etc. In the final analysis, it can be reflected in two aspects: the diversity of the target or the target itself, such as skin color, emotion , occlusion, posture, etc.; the diversity of shooting conditions other than the target, such as light, backlight, front light, exposure level, noise of shooting equipment, and the quality of anti-shake function, etc., are reflected in the image, that is, the contrast and clarity of the image , signal-to-noise ratio, detail restoration, etc. Therefore, the purpose of this program is to make quantitative judgments on the degree of influence of different types of diversity on the algorithm, and then based on the judgment results, solve the problems that cannot be achieved by a single recognition rate:
[0070] 1) Multi-dimensional index evaluation;
[0071] 2) Problem dimension analysis...
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 a face recognition algorithm evaluation method based on quality dimensions. The method comprises a face recognition evaluation method based on target correlation and a face recognition evaluation method based on non-target correlation, and for a single face recognition algorithm, the sensitivity of the algorithm to various different parameters can be obtained through multi-dimensional evaluation, so that algorithm optimization is performed for the different parameters; for different face recognition algorithms, a more detailed comparison result can be provided, and an optimal recognition algorithm is given in combination with an application environment.
Description
technical field [0001] The invention relates to the technical field of image processing. Background technique [0002] Face recognition algorithms have great reference value in current authorized applications and academic research. There are more and more face recognition algorithms proposed based on different starting points and principles. However, the relationship between image quality and recognition algorithms only exists in Refuse to recognize the direction, that is, assume that the quality of the provided image is too low, then re-acquire without recognition. In fact, based on the quality system, it is easier to know the pros and cons of the algorithm, it can provide more objective algorithm evaluation indicators, and at the same time provide the direction of optimizing the algorithm. [0003] In terms of face recognition algorithm evaluation, it mainly focuses on recognition rate and operating efficiency, including: [0004] 1) An evaluation model based on the reco...
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): G06K9/03G06K9/00
CPCG06V40/16G06V40/174G06V40/172G06V10/993
Inventor 董波王道宁张亚东陶亮廖志梁
Owner 易诚高科(大连)科技有限公司