Unlock instant, AI-driven research and patent intelligence for your innovation.

Video frame feature extraction model training method, system and device and storage medium

A technology of feature extraction and training methods, applied in neural learning methods, biological neural network models, character and pattern recognition, etc., can solve problems such as inability to adapt to image transformation, noise interference, high missed detection rate, etc., to increase transformation adaptability, The effect of reducing the missed detection rate and improving the effect

Pending Publication Date: 2021-11-05
人民中科(北京)智能技术有限公司
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The invention provides a training method, system, equipment and storage medium for a video frame feature extraction model, which are used to solve the problems in the prior art that are easily disturbed by noise, cannot adapt to image transformation, and have a high rate of missed detection

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Video frame feature extraction model training method, system and device and storage medium
  • Video frame feature extraction model training method, system and device and storage medium
  • Video frame feature extraction model training method, system and device and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] Embodiment 1, the training method of the video frame feature extraction model of the present embodiment, see figure 1 shown, including the following main steps:

[0031] S110. After image enhancement is performed on any batch of original video frame images, two channels of enhanced video frame image sets are acquired.

[0032] Specifically, the original video frame data set is collected first, for example, about 600,000 original video frame images are collected, and it is ensured that there are not a large number of identical original video frame images in the data set. Specifically, a random image data set crawled from the web can be used. Although the same original video frame images cannot be avoided, the probability is small, so it can satisfy that there are not many identical original video frame images in the data set; or use an existing data set. Then image enhancement is performed on any batch of original video frame images in the original video frame data set....

Embodiment 2

[0051] Embodiment 2, the training system of the video frame feature extraction model of the present embodiment, see Figure 7 As shown, it includes: an image enhancement unit 210 , a two-way feature extraction unit 220 , a similarity matrix unit 230 , a loss value calculation unit 240 , a judging unit 250 , and a one-way network extraction unit 260 .

[0052] The image enhancement unit 210 is configured to obtain two channels of enhanced video frame image sets after any batch of original video frame images undergoes image enhancement. Specifically, the original video frame data set is collected first, for example, about 600,000 original video frame images are collected, and it is ensured that there are not a large number of identical original video frame images in the data set. Specifically, a random image data set crawled from the web can be used. Although the same original video frame images cannot be avoided, the probability is small, so it can be satisfied that there are n...

Embodiment 3

[0059] Embodiment 3, the computer equipment of this embodiment, see Figure 8 As shown, the displayed computer device 300 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present invention.

[0060] Such as Figure 8 As shown, computer device 300 takes the form of a general-purpose computing device. Components of computer device 300 may include, but are not limited to: one or more processors or processing units 301 , system memory 302 , bus 303 connecting different system components (including system memory 302 and processing unit 301 ).

[0061] Bus 303 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus structures. These architectures include, by way of example, but are not limited to Industry Standard Architecture (ISA) bus, Micro Channel Architectur...

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

No PUM Login to View More

Abstract

The invention discloses a video frame feature extraction model training method, system and device and a storage medium. The method comprises the following steps: S1, performing image enhancement on any batch of original video frame images to obtain two paths of enhanced video frame image sets; S2, respectively inputting the two paths of enhanced video frame image sets into two paths of feature extraction sub-networks included in a comparison training network to obtain two paths of feature sets; S3, performing cross-correlation multiplication calculation on the two paths of feature sets along feature dimensions to obtain a similarity matrix; S4, calculating a loss value of the similarity matrix; S5, if the loss value is greater than the threshold value, adjusting the comparison training network according to the loss value, and returning to S1, otherwise, judging comparison training network fitting, and turning to S6; and S6, extracting any feature extraction sub-network included in the comparison training network, and completing training of the video frame feature extraction model. The system comprises an image enhancement unit, a double-path feature extraction unit, a similarity matrix unit, a loss value calculation unit, a judgment unit and a single-path network extraction unit.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a training method, system, device and storage medium for a video frame feature extraction model. Background technique [0002] Image sample comparison, also known as image homologous comparison, refers to an image processed by scaling, compression, partial rotation, color transformation, format conversion, partial cropping, mosaic, blurring, labeling, text occlusion, etc. Still correctly matches the original image. [0003] The current image sample extraction method is basically based on the traditional manual feature extraction method, including directly extracting Locality Sensitive Hash (LSH) to build an index, or extracting color distribution features, HOG, SIFT and other gradient features for feature extraction. This type of method is easily disturbed by noise and cannot adapt to various transformation operations of pictures. For example, methods based on color dist...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/214
Inventor 张朝王坚李兵余昊楠胡卫明
Owner 人民中科(北京)智能技术有限公司