Small sample target detection method based on attention and contrast learning
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology of target detection and attention, applied in the field of image processing and artificial intelligence, can solve problems such as insufficient utilization, different parameter settings, knowledge transfer and forgetting, etc.
Pending Publication Date: 2021-09-14
KUNMING UNIV OF SCI & TECH
View PDF0 Cites 9 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0009] The meta-learning method is actually an attention mechanism, which is based on the attention of the support set and the query set, but it does not make full use of the relationship between instances. The attention pays more attention to the differences between different instances. Contrastive learning
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0066] Embodiment 1: as Figure 1-6 As shown, a small-sample object detection method based on attention and contrastive learning is characterized in that it includes a small-sample mosaic data enhancement module, a coordinate compensation attention region proposal network, a new and old class discrimination module, and a comparative learning module.
[0067] Small sample mosaic data enhancement module, including sample selection strategy and mosaic data enhancement method;
[0068] Coordinate compensation attention region proposal network, including region proposal network, regression coordinate compensation attention branch, coordinate compensation loss;
[0069] New and old class discrimination module, including feature extraction layer and class discrimination layer;
[0070] Contrastive learning module, including feature map layer, contrastive loss and joint training strategy;
[0071] On the basis of the general two-stage object detection baseline R-CNN, the above modul...
Embodiment 2
[0122] Embodiment 2: This embodiment describes the present invention in combination with specific data.
[0123] The model training is divided into two stages: the first stage is base class training; the second stage is the fine-tuning stage for new classes with small samples. Among them, the number of training samples of the base class is sufficient, and the training data of the new class is divided into 1, 3, 5 and 10 samples for each class according to the task. The model is trained and verified on the COCO dataset and the VOC dataset respectively. The COCO dataset has 80 categories, of which 60 are base categories and 20 are new categories; the VOC dataset has 20 categories, of which 15 1 class is the base class and 5 classes are new classes.
[0124] Hyperparameter settings: Gradient descent uses standard SGD, momentum is 0.9, weight decay is e -4 , batchsize is 16. The software environment is Ubuntu18.04, Cuda10.2, Pytorch1.4.0, Python3.6; the hardware environment is 8...
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 relates to a small sample target detection method based on attention and contrast learning, and belongs to the field of artificial intelligence and image processing. The invention relates to the small sample target detection method which combines data enhancement, an attention region suggestion network (Attention RPN) and comparative learning. The method is based on a Faster R-CNN (Convolutional Neural Network) network, and comprises the following steps of: adopting a Few-shot Mosaic data enhancement module for enriching the comparison between a small sample background instance and a Novelclass instance and a Base class instance, enhancing the attention of a model on a foreground by an Attention RPN (Regression Coordinate Compensation) module based on regression coordinate compensation, and improving the expression of instance-level features by a contrast learning module. According to the method, the new class detection precision of the Faster R-CNN on a small sample is improved, and meanwhile, relatively high base class detection precision is kept; the dependency of Faster R-CNN on the new class training sample size is reduced, the new class migration ability is improved, and the effectiveness of the method is verified on COCO and VOC data sets.
Description
technical field [0001] The invention relates to a small-sample target detection method, specifically designs a small-sample target detection method based on attention and contrastive learning, and belongs to the fields of artificial intelligence and image processing. Background technique [0002] Target detection is to detect the target and calibrate the target position in the picture at the same time, so the task of target detection is more complicated than that of picture classification, involving not only classification tasks but also regression tasks. Since the data samples have long-tail distribution characteristics on the one hand, and some samples are difficult to obtain on the other hand, the cost of labeling is huge. Therefore, there is also a small sample problem in target detection, that is, after training on a base class sample with a large sample size, fine-tuning training is performed on a new class with a small sample size and a small sample size, so that 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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to view more 
Login to view more