CN116029968A - Monkey pox infection skin image detection method and device, electronic equipment and storage medium - Google Patents
Monkey pox infection skin image detection method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116029968A CN116029968A CN202211101846.5A CN202211101846A CN116029968A CN 116029968 A CN116029968 A CN 116029968A CN 202211101846 A CN202211101846 A CN 202211101846A CN 116029968 A CN116029968 A CN 116029968A
- Authority
- CN
- China
- Prior art keywords
- training
- image
- probability model
- loss function
- denoising
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 208000005871 monkeypox Diseases 0.000 title claims abstract description 128
- 238000001514 detection method Methods 0.000 title claims abstract description 50
- 208000015181 infectious disease Diseases 0.000 title description 18
- 238000009792 diffusion process Methods 0.000 claims abstract description 116
- 238000000034 method Methods 0.000 claims abstract description 103
- 238000005070 sampling Methods 0.000 claims abstract description 33
- 238000012549 training Methods 0.000 claims description 151
- 230000006870 function Effects 0.000 claims description 118
- 238000012545 processing Methods 0.000 claims description 61
- 230000008569 process Effects 0.000 claims description 38
- 238000012795 verification Methods 0.000 claims description 35
- 238000012360 testing method Methods 0.000 claims description 21
- 230000036074 healthy skin Effects 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000011478 gradient descent method Methods 0.000 claims description 5
- 238000013473 artificial intelligence Methods 0.000 abstract description 17
- 238000005516 engineering process Methods 0.000 description 22
- 238000010801 machine learning Methods 0.000 description 14
- 238000004891 communication Methods 0.000 description 11
- 238000003058 natural language processing Methods 0.000 description 6
- 238000003745 diagnosis Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000011160 research Methods 0.000 description 4
- 238000012935 Averaging Methods 0.000 description 3
- 241000700627 Monkeypox virus Species 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000013145 classification model Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 241000282412 Homo Species 0.000 description 1
- 241000700629 Orthopoxvirus Species 0.000 description 1
- 241000700647 Variola virus Species 0.000 description 1
- 208000025259 Viral Zoonoses Diseases 0.000 description 1
- 241000700605 Viruses Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003042 antagnostic effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000005180 public health Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Abstract
The embodiment of the application provides a monkey pox infected skin image detection method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: obtaining a skin image to be detected, and carrying out noise adding treatment on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding treatment result; obtaining a sampling result from standard normal distribution; then calculating to obtain a loss function value; and judging whether the loss function value belongs to a first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image. According to the embodiment of the application, the monkey pox infected skin image can be accurately detected, and the detection accuracy is high.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a monkey pox infected skin image detection method and device, electronic equipment and a storage medium.
Background
Monkey pox is a viral zoonotic disease that can be transmitted to humans by intimate contact and belongs to the genus orthopoxvirus as smallpox eliminated in 1980. By day 30 of 5 months 2022, about 30 countries worldwide report 257 cases of conchopox diagnosis, with about 120 suspected cases. The world health organization has rated the global public health risk of monkey pox as moderate, and is extremely critical for the general public to protect and diagnose and treat in advance.
At present, a model for detecting potential monkey pox infection skin images is mainly based on a deep neural network framework, corresponding labels are added to existing monkey pox infection skin images and healthy skin images, a supervised classification model is trained, and detection and diagnosis of the condition of the monkey pox are achieved during testing.
However, current methods of monkey pox infection skin image diagnosis have significant drawbacks. The supervised model needs to label the monkey pox infected skin image and the healthy skin image data, and needs a great deal of time cost and personnel investment, so that unnecessary burden is brought to the experiment. Meanwhile, the training stability of the classification model based on the countermeasure thought is very challenging, a way for realizing Nash balance is difficult to find in the countermeasure training process, the situation that the loss function is continuous and vibration is difficult to converge easily occurs, effective excavation of monkey pox infection image data characteristics is difficult to complete, and a detection task cannot be completed efficiently and accurately.
Disclosure of Invention
The embodiment of the application mainly aims to provide a method and a device for detecting a monkey pox infected skin image, electronic equipment and a storage medium, and aims to efficiently and accurately detect the monkey pox infected skin image.
To achieve the above object, a first aspect of embodiments of the present application provides a method for detecting a monkey pox infection skin image, the method comprising:
acquiring a skin image to be detected;
carrying out noise adding treatment on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding treatment result;
obtaining a sampling result from standard normal distribution;
calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
judging whether the loss function value belongs to the first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
In some embodiments, the step of calculating a loss function value according to the skin image to be detected, the noise processing result and the sampling result is performed by the following loss function formula:
wherein ,
where Loss represents the Loss function, ε represents the sampling result obtained from the normal distribution, ε θ Representing the diffusion denoising probability model trained in advance, x 0 Representing the skin image to be detected, beta t To represent the noise variance of the T-th step, an equal-ratio array between 0 and 1 is taken, t=1, 2,3 … T, and T is the number of noise adding steps.
In some embodiments, the first threshold interval is determined by:
acquiring a training image dataset comprising a plurality of monkey pox infected skin images;
carrying out noise adding treatment on each image in a training image data set through the diffusion denoising probability model trained in advance, and calculating to obtain an average loss function value;
and determining a first threshold interval according to the average loss function value and the fluctuation proportion threshold.
In some embodiments, the step of performing noise adding processing on each image in the training image data set through the pre-trained diffusion denoising probability model and calculating to obtain an average loss function value includes:
performing standardization processing on pixel values of all images in the training image data set;
carrying out noise adding processing on all the standardized processed images through the diffusion denoising probability model trained in advance to obtain a noise adding processing result set, wherein the noise adding processing result set comprises noise adding processing results corresponding to each image in the training image data set;
According to the noise processing result set and the standard normal distribution sampling result set, calculating a loss function value set by using the loss function, wherein the loss function value set comprises a loss function value corresponding to each image in the training image data set;
and carrying out average calculation on each loss function value in the loss function value set to obtain an average loss function value.
In some embodiments, the method further includes pre-training the diffuse denoising probability model, specifically comprising:
constructing the training image dataset;
performing standardization processing on all image pixel values in the training image data set;
acquiring a monkey pox infected skin image in the training image data set for training;
according to the loss function, updating parameters of the diffusion denoising probability model by using a random gradient descent method;
acquiring a next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image data set are trained completely, and completing one iteration;
judging whether the diffusion denoising probability model training meets the ending condition or not, outputting the diffusion denoising probability model if the ending condition is met, and continuing training if the ending condition is not met.
In some embodiments, the obtaining the next monkey pox infected skin image according to the updated parameters of the diffusion denoising probability model for training until all the monkey pox infected skin images in the training image dataset are trained, and after completing one iteration, the pre-training the diffusion denoising probability model further includes:
obtaining a verification image dataset comprising a plurality of monkey pox infected skin images and a plurality of healthy skin images;
inputting the verification image data set into the diffusion denoising probability model to test the detection accuracy of the diffusion denoising probability model;
and obtaining a test result, and controlling the training process of the diffusion denoising probability model by using an early-stop method according to the test result.
In some embodiments, the step of determining whether the diffuse denoising probability model training meets an end condition, outputting the diffuse denoising probability model if the diffuse denoising probability model meets the end condition, and continuing training if the diffuse denoising probability model does not meet the end condition specifically includes:
acquiring a verification error of the verification image dataset;
outputting the diffusion denoising probability model when the verification error continuously rises in the training process of the preset iteration times;
And when the verification error does not continuously rise in the training process of the preset iteration times, continuing to train the diffusion denoising probability model.
To achieve the above object, a second aspect of the embodiments of the present application provides a device for detecting a monkey pox infected skin image, the device comprising:
the first acquisition module is used for acquiring a skin image to be detected;
the processing module is used for carrying out noise adding processing on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding processing result;
the second acquisition module is used for acquiring sampling results from the standard normal distribution;
the calculation module is used for calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
the judging module is used for judging whether the loss function value belongs to the first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
To achieve the above object, a third aspect of the embodiments of the present application proposes an electronic device comprising a memory and a processor, a program stored on the memory and executable on the processor, and a data bus for enabling a connection communication between the processor and the memory, the program, when executed by the processor, implementing the method according to the first aspect.
To achieve the above object, a fourth aspect of the embodiments of the present application proposes a storage medium, which is a computer-readable storage medium, for computer-readable storage, the storage medium storing one or more programs executable by one or more processors to implement the method described in the first aspect.
According to the method and the device for detecting the monkey pox infected skin image, the electronic equipment and the storage medium, the skin image to be detected is obtained, and then the noise adding treatment is carried out on the skin image to be detected through a pre-trained diffusion denoising probability model, so that a noise adding treatment result is obtained; obtaining a sampling result from the standard normal distribution, and calculating to obtain a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result; and finally judging whether the loss function value belongs to a first threshold value interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image. The detection is carried out through the pre-trained diffusion denoising probability model, the appearance information of the monkey pox virus in human morbidity is fully considered, the detection can be completed under the condition that the image data do not need to be marked, the detection performance is better, whether the skin image to be detected is a monkey pox infected skin image or not can be detected efficiently and accurately, and the detection accuracy is high.
Drawings
FIG. 1 is an application scenario diagram of a diffuse denoising probabilistic model training method in one embodiment;
FIG. 2 is a flow chart of a method for training a diffuse denoising probability model according to one embodiment;
fig. 3 is a flowchart of step S206 in fig. 2;
fig. 4 is a flowchart of a method for detecting a monkey pox infected skin image provided in an embodiment of the present application;
FIG. 5 is a flowchart of steps for determining a first threshold interval provided by an embodiment of the present application;
fig. 6 is a flowchart of step S502 in fig. 5;
fig. 7 is a schematic structural diagram of a device for detecting a monkey pox infected skin image according to an embodiment of the present application;
fig. 8 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
It should be noted that although functional block division is performed in a device diagram and a logic sequence is shown in a flowchart, in some cases, the steps shown or described may be performed in a different order than the block division in the device, or in the flowchart. The terms first, second and the like in the description and in the claims and in the above-described figures, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the present application.
First, several nouns referred to in this application are parsed:
artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.
The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning, automatic driving, intelligent traffic and other directions.
Computer Vision (CV) is a science of how to make a machine "look at", and more specifically, to replace human eyes with a camera and a Computer to perform machine Vision such as recognition, tracking and measurement on a target, and further perform graphic processing, so that the Computer processes the target into an image more suitable for human eyes to observe or transmit to an instrument to detect. As a scientific discipline, computer vision research-related theory and technology has attempted to build artificial intelligence systems that can acquire information from images or multidimensional data. Computer vision technologies typically include image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content/behavior recognition, three-dimensional object reconstruction, 3D technology, virtual reality, augmented reality, synchronous positioning and mapping, autopilot, intelligent transportation, etc., as well as common biometric technologies such as face recognition, fingerprint recognition, etc.
Natural language processing (Nature Language processing, NLP) is an important direction in the fields of computer science and artificial intelligence. It is studying various theories and methods that enable effective communication between a person and a computer in natural language. Natural language processing is a science that integrates linguistics, computer science, and mathematics. Thus, the research in this field will involve natural language, i.e. language that people use daily, so it has a close relationship with the research in linguistics. Natural language processing techniques typically include text processing, semantic understanding, machine translation, robotic questions and answers, knowledge graph techniques, and the like.
Machine Learning (ML) is a multi-domain interdisciplinary, involving multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is specially studied how a computer simulates or implements learning behavior of a human to acquire new knowledge or skills, and reorganizes existing knowledge structures to continuously improve own performance. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, confidence networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like.
The scheme provided by the embodiment of the application relates to the technologies of artificial intelligence, such as computer vision technology, natural language processing, machine learning and the like, and is specifically described by the following embodiments:
with the rapid development of science and technology, artificial intelligence technology is getting more and more attention. In particular, machine learning techniques in artificial intelligence are widely used. Machine learning techniques can typically be used to train machine learning models capable of detecting images of monkey pox infected skin. For example, using the monkey pox infected skin image as a training sample, a detection model is trained using machine learning techniques to detect the monkey pox infected skin image.
At present, a model for detecting potential monkey pox infection skin images is mainly based on a deep neural network framework, corresponding labels are added to existing monkey pox infection skin images and healthy skin images, a supervised classification model is trained, and detection and diagnosis of the condition of the monkey pox are achieved during testing.
Supervised learning is a traditional model training method. However, for the supervised learning method, a large amount of labeling data needs to be manually added to the sample, which results in a relatively high model training cost.
In view of this, the embodiment of the application proposes a diffusion denoising probability model to detect the monkey pox infected skin image. The method fully considers the expression information of the monkey pox virus in human morbidity, has better virus information detection performance under the condition that image data do not need to be marked, and can efficiently complete diagnosis tasks. Meanwhile, the detection accuracy of the unsupervised model is tested by verifying the image data set, so that the training condition and the detection capability of the model can be intuitively known. Meanwhile, the diffusion denoising probability model avoids an antagonistic model frame and a training mechanism, the training process is stable, the performance is more stable, and the detection result of the monkey pox infected skin image is more reliable.
Fig. 1 is an application scenario diagram of a diffuse denoising probabilistic model training method in one embodiment. Referring to fig. 1, the application scenario includes a server 110 and a terminal 120 connected through a network. The terminal 120 may be a medical device, a desktop computer, or a mobile terminal. The medical device is a terminal device capable of acquiring an image. The mobile terminal may include at least one of a cell phone, a tablet computer, a notebook computer, a personal digital assistant, a wearable device, and the like. The server 110 may be implemented as a stand-alone server or as a server cluster of multiple physical servers. It will be appreciated that in other embodiments, the server 110 may be replaced with a terminal having the ability to perform the diffuse denoising probability model training method of embodiments of the present application.
The terminal 120 may perform image acquisition and send the acquired monkey pox infected skin image as a training image dataset to the server 110 to provide the server 110 with a training image dataset capable of model training. For example, in a medical scenario, the medical device may collect a skin image of the infection with the monkey pox and provide the skin image to the server 110, and the server 110 may perform machine learning training with the skin image of the infection with the monkey pox as a training image data set to train a detection model capable of identifying the skin image of the infection with the monkey pox. It will be appreciated that the server 110 may also directly obtain stored training image data sets transmitted by the terminal 120.
It will be appreciated that the server 110 needs to train the diffuse denoising probability model through multiple rounds of iterations. Therefore, in each iterative training process, the server 110 may input one monkey pox infected skin image in the training image data set into the diffusion denoising probability model for training, and then update the parameters of the diffusion denoising probability model by using the random gradient descent method according to the loss function. And acquiring a next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image data set are completely trained, and completing one iteration. And finally judging whether the diffuse denoising probability model training meets the end condition, outputting the diffuse denoising probability model if the diffuse denoising probability model training meets the end condition, and continuing training if the diffuse denoising probability model training does not meet the end condition.
It will be appreciated that the terminal 120 may also collect images of the monkey pox infected skin and images of the healthy skin and send the collected images of the monkey pox infected skin and images of the healthy skin as a verification image dataset to the server 110 to provide the server 110 with a verification image dataset enabling model verification. The server 110 may acquire a verification image dataset and then input the verification image dataset into the diffusion denoising probability model to test the detection accuracy of the diffusion denoising probability model; and then obtaining a test result, and controlling the training process of the diffusion denoising probability model by using an early-stop method according to the test result. When the server 110 detects that the verification error continuously rises in the training process of the preset iteration times, the training is completed, and at the moment, a diffusion denoising probability model is output;
It can be appreciated that the method for training the diffusion denoising probability model in the embodiments of the present application is equivalent to training a machine learning model capable of identifying the skin image of the monkey pox infection by using an artificial intelligence technique. The training method of the diffusion denoising probability model in each embodiment of the application is equivalent to using a machine learning technology.
Fig. 2 is a flow chart of a training method of a diffuse denoising probability model in one embodiment, which can be applied to a computer device, and the computer device is mainly taken as the server 110 in fig. 1 for illustration. Referring to fig. 2, the method specifically includes steps S201 to S206:
step S201, constructing a training image data set, wherein the training image data set comprises a plurality of monkey pox infected skin images;
step S202, carrying out standardization processing on all image pixel values in the training image data set;
step S203, obtaining a monkey pox infected skin image in a training image data set for training;
step S204, updating parameters of the diffusion denoising probability model by using a random gradient descent method according to the loss function;
step S205, acquiring a next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image data set are trained, and completing one iteration;
Step S206, judging whether the diffuse denoising probability model training meets the end condition, outputting the diffuse denoising probability model if the diffuse denoising probability model training meets the end condition, and continuing training if the diffuse denoising probability model training does not meet the end condition.
In the embodiment of the application, only the monkey pox infected skin image dataset is required to be constructed as the training image dataset, and the training image dataset is input into the diffusion denoising probability model for training. Training images refer to images used for model training. The image may be a picture or a video frame in a video. It should be noted that the pixel values of all images in the training image data set need to be normalized, for example, the pixel values of all images in the training image data set are normalized to the interval [ -1,1]And then inputting the model into a diffusion denoising probability model for training. The operation of the diffusion denoising probability model is divided into a forward denoising process and a reverse denoising process. In the noise adding process, the image x is input first 0 Each subsequent step adds Gaussian noise to the previous stepThe sound eventually becomes almost pure gaussian noise. The denoising process is opposite to the denoising process, a noise sample is obtained by sampling from the standard normal Ethernet distribution, the noise is further denoised step by step, and finally, a sample in the image data distribution is obtained. Specifically, during the noise adding process, input x 0 The image in the monkey pox skin image data set is subjected to a 1-step noise adding diffusion process
Obtaining x 1 Then the noise adding diffusion process is carried out in step 1>
Obtaining x 2 And so on, the noise-added diffusion process q (x) with Markov performance is carried out through the T steps T |x T-1 ) Obtaining x T ，x T To obey the gaussian distribution with a mean of 0, the variance is the unit array I. Wherein I is an identity matrix, beta t Is a constant representing the noise variance of the first step, and typically takes the form of an equal-ratio array of values between 0 and 1, t=1, 2,3 … T, T being the number of noise steps. In the denoising process, x is the sum of T The distributed samples are subjected to a 1-step denoising process p (x T-1 |x T ) Obtaining x T-1 Then through 1 step denoising process p (x) T-2 |x T-1 ) Obtaining x T-2 And so on, the denoising process p (x) with Markov performance is carried out through T steps T-1 |x T ) Restoring the original data image to obtain x 0 . The model training loss function is the distance between the noise vector of the standard normal distribution and the result of the noise adding of the input data, and when the loss is smaller, the final distribution forward standard normal distribution of the noise adding diffusion process is gradually close to or even overlapped.
In the training process, firstly, inputting a monkey pox infected skin image into a diffusion denoising probability model to perform T-step denoising treatment, randomly taking a denoising result of a time step T from the steps 1-T, then obtaining a sampling result corresponding to the time step T from standard normal distribution, calculating a loss function value according to a loss function value, and updating parameters of the diffusion denoising probability model by using a random gradient descent method according to the loss function value; and acquiring a next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image data set are completely trained, and completing one iteration. Wherein, the loss function expression is:
wherein ,
where Loss represents the Loss function, ε represents the sampling result obtained from the normal distribution, ε θ Representing the diffusion denoising probability model trained in advance, x 0 Representing an image of monkey pox infected skin, beta t To represent the noise variance of the T-th step, an equal-ratio array between 0 and 1 is taken, t=1, 2,3 … T, and T is the number of noise adding steps.
In this embodiment of the present application, after completing one iteration, the detection accuracy of the diffusion denoising probability model by the input verification image dataset is tested, and specifically includes the following steps:
step S205-1, acquiring a verification image data set, wherein the verification image data set comprises a plurality of monkey pox infected skin images and a plurality of healthy skin images;
step S205-2, inputting the verification image data set into a diffusion denoising probability model to test the detection accuracy of the diffusion denoising probability model;
and step S205-3, obtaining a test result, and controlling the training process of the diffusion denoising probability model by using an early-stop method according to the test result.
In the embodiment of the application, after one iteration is completed, all images in a training image dataset are input into a diffusion denoising probability model, a result after diffusion denoising is calculated, all loss function values corresponding to all images are calculated through a loss function, and an average loss function value is obtained by averaging 
Specifically, since the training image dataset contains a plurality of monkey pox infected skin images, a monkey pox infected skin image can be input first, a noise adding result is output after noise adding, the noise adding result corresponding to the time step t in the noise adding process is randomly selected, the sampling result corresponding to the time step t is obtained from the standard normal distribution, and the loss function value is calculated by using the loss function. Then, inputting the next monkey pox infected skin image, similarly calculating the corresponding loss function value, and similarly, calculating all the loss function values corresponding to all the monkey pox infected skin images in the training image data set, and then carrying out average calculation on all the loss function values to obtain the average loss function value. For example, the training image data set includes N monkey pox infected skin images, the first monkey pox infected skin image in the training image data set is input into the trained diffusion denoising probability model, and the calculated corresponding loss function value is L 1 The method comprises the steps of carrying out a first treatment on the surface of the Then inputting a second monkey pox infected skin image in the training image data set into the trained diffusion denoising probability model, and calculating the corresponding loss function value as L 2 The method comprises the steps of carrying out a first treatment on the surface of the In this way, the Nth monkey pox infected skin image in the training image data set is input into the trained diffusion denoising probability model, and the calculated corresponding loss function value is L N . Then, the average loss function value is calculated to be +.>
Then a fluctuation ratio threshold α is set, such as α=0.1. Then a first threshold interval is determined based on the average loss function value and the fluctuation ratio threshold, e.g. about to +.>
As a first threshold interval for detecting monkey pox infected skin images. And then respectively inputting the monkey pox infected skin image and the healthy skin image to test the detection accuracy of the diffusion denoising probability model. For example, when the input is a monkey pox infected skin image, the input of a diffusion denoising probability model is obtainedThe result is obtained, then the loss function value is calculated by using the loss function, and then whether the loss function value belongs to
If the detection result is correct, the detection result is incorrect. When the input is a healthy skin image, an output result of a diffusion denoising probability model is obtained, then a loss function value is calculated by using a loss function, and then whether the loss function value belongs to +.>
If the detection result is not correct, the detection result is correct. Through the test mode, the early stop method is used for controlling the training process of the diffusion denoising probability model. The detection accuracy of the diffusion denoising probability model can be guaranteed through testing the detection accuracy of the diffusion denoising probability model. The training process of the diffusion denoising probability model is controlled by the early stop method, so that the training time of the diffusion denoising probability model can be reduced, and the training efficiency is improved.
Illustratively, testing the detection accuracy of the diffusion denoising probability model by verifying the image dataset may be performed by:
a verification image dataset is obtained, the verification image dataset comprising 100 monkey pox infected skin images and 100 healthy skin images. Firstly, 100 monkey pox infection skin images are sequentially input into a diffusion denoising probability model, loss function values corresponding to the 100 monkey pox infection skin images are calculated respectively, whether the loss function values corresponding to the 100 monkey pox infection skin images are in a first threshold interval or not is sequentially judged, if the loss function values corresponding to the 80 monkey pox infection skin images are found to be in the first threshold interval through judgment, the loss function values corresponding to the 20 monkey pox infection skin images are not in the first threshold interval, and the first accuracy is recorded to be 80%. And sequentially inputting 100 healthy skin images into a diffusion denoising probability model, respectively calculating the loss function values corresponding to the 100 healthy skin images, sequentially judging whether the loss function values corresponding to the 100 healthy skin images are in a first threshold interval, if so, finding that the loss function values corresponding to the 10 healthy skin images are in the first threshold interval, and if not, recording that the second accuracy is 90%. And calculating the average accuracy rate to be 85% according to the first accuracy rate and the second accuracy rate.
In the embodiment of the application, after the detection accuracy of the diffusion denoising probability model is obtained through verification of the image dataset test, whether the detection accuracy of the diffusion denoising probability model meets the preset requirement is further judged. For example, the preset accuracy rate is required to be more than 95%, and the detection accuracy rate of the diffusion denoising probability model obtained through the verification image dataset test is 85%, which indicates that the detection accuracy rate of the diffusion denoising probability model does not reach the standard, and at the moment, iterative training needs to be continued. If the detection accuracy of the diffusion denoising probability model obtained through the verification image data set test is 98%, the detection accuracy of the diffusion denoising probability model is up to the standard, at the moment, interpretation training is performed, and the trained diffusion denoising probability model is output.
Referring to fig. 3, in some embodiments, step S206 may include, but is not limited to, steps S301 to S303:
step S301, acquiring verification errors of a verification image dataset;
step S302, when the verification error continuously rises in the training process of the preset iteration times, outputting a diffusion denoising probability model;
in step S303, when the verification error does not continuously rise in the training process of the predetermined number of iterations, the diffusion denoising probability model is continuously trained.
In the embodiment of the application, if the verification error continuously rises in the training process of the preset iteration times, for example, the verification error continuously rises in the iterative training process of more than 10 times, the detection accuracy of the model is proved to reach the requirement, and at the moment, the training is completed, and a trained diffusion denoising probability model can be output. If the verification error does not continuously rise in the training process of the preset iteration times, for example, the verification error does not continuously rise in the iterative training process of more than 10 times, the detection accuracy of the model does not meet the requirement yet, and the diffusion denoising probability model needs to be continuously trained.
It should be noted that, in the embodiments of the present application, the method for training the diffusion denoising probability model is not directly applied to the living target object in the real world, but rather performs model training by collecting an image of the monkey pox infected skin as a training image data set, which belongs to machine learning training combined with an image processing technology.
It should be noted that model training using a training image dataset belongs to unsupervised learning. The method does not need to spend a great deal of manpower cost to mark the monkey pox infected skin image and the healthy skin image, can deeply mine global information in an unsupervised state, and efficiently completes the detection task of the monkey pox infected skin image; meanwhile, a diffusion denoising probability model is adopted, so that the training process is more stable and the performance is more stable.
In the embodiment of the application, after the diffusion denoising probability model is trained, the skin image to be detected can be detected by using the trained diffusion denoising probability model. Fig. 4 is an alternative flowchart of a method for detecting a monkey pox infected skin image provided in an embodiment of the present application, where the method in fig. 4 may include, but is not limited to, steps S401 to S405.
Step S401, obtaining a skin image to be detected;
step S402, carrying out noise adding processing on a skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding processing result;
step S403, obtaining a sampling result from the standard normal distribution;
step S404, calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
step S405, judging whether the loss function value belongs to a first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
In the embodiment of the present application, the method is shown in fig. 2After the training method of the model is trained, the trained model can be further utilized to detect the skin image to be detected. Specifically, firstly, a skin image to be detected is obtained, the skin image to be detected is input into a pre-trained diffusion denoising probability model for denoising, and a denoising result is obtained. And then obtaining a sampling result from the standard normal distribution, and calculating to obtain a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result. Then inputting all images in the training image data set into a diffusion denoising probability model, calculating a diffusion denoising result, calculating all loss function values corresponding to all images through a loss function, and averaging to obtain an average loss function value 
Setting a fluctuation ratio threshold α, e.g., α=0.1, will +.>
As a first threshold interval for detecting monkey pox infected skin images. And finally judging whether the loss function value belongs to a first threshold value interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image. Whether the skin image to be detected is a monkey pox infected skin image or not can be detected efficiently and accurately, and the detection accuracy is high.
It should be noted that, in this embodiment, after the skin image to be detected is obtained, the pixels of the skin image to be detected need to be subjected to value normalization processing. Illustratively, the pixel values of the skin image to be detected are normalized to the interval [ -1,1], and then input into a pre-trained diffusion denoising probability model.
In the embodiment of the present application, step S404 is performed by the following loss function formula:
wherein ,
Where Loss represents the Loss function, ε represents the sampling result obtained from the normal distribution, ε θ Representing the diffusion denoising probability model trained in advance, x 0 Representing the skin image to be detected, beta t To represent the noise variance of the T-th step, an equal-ratio array between 0 and 1 is taken, t=1, 2,3 … T, and T is the number of noise adding steps.
Referring to fig. 5, in some embodiments, the first threshold interval is determined by steps S501 to S503:
step S501, a training image data set is obtained, wherein the training image data set comprises a plurality of monkey pox infected skin images;
step S502, carrying out noise adding processing on each image in a training image data set through a pre-trained diffusion denoising probability model, and calculating to obtain an average loss function value;
step S503, determining a first threshold interval according to the average loss function value and the fluctuation proportion threshold.
In the embodiment of the application, a training image data set is acquired, and similarly, the pixel values of all images in the training image data set are normalized to the interval [ -1,1]Inputting the result into a pre-trained diffusion denoising probability model, calculating the diffusion denoising result, calculating all loss function values corresponding to all images through a loss function, and averaging to obtain an average loss function value
Setting a fluctuation ratio threshold α, e.g., α=0.1, will +.>
As a first threshold interval for detecting monkey pox infected skin images.
Referring to fig. 6, in some embodiments, step S502 may include, but is not limited to, steps S601 to S604:
step S601, carrying out standardization processing on pixel values of all images in a training image data set;
Step S602, carrying out noise adding processing on all the standardized processed images through a pre-trained diffusion denoising probability model to obtain a noise adding processing result set, wherein the noise adding processing result set comprises noise adding processing results corresponding to each image in training image data;
step S603, calculating a loss function value set by using a loss function according to the noise processing result set and the sampling result set of standard normal distribution, wherein the loss function value set comprises a loss function value corresponding to each image in the training image data set;
step S604, performing an average calculation on each loss function value in the set of loss function values to obtain an average loss function value.
In the embodiment of the application, after the diffusion denoising probability model is trained, the training image dataset is further input into the trained diffusion denoising probability model. Specifically, the pixel values of all images in the training image data set are normalized, for example, to the interval [ -1,1]. Then, carrying out noise adding processing on all the standardized processed images through a pre-trained diffusion denoising probability model to obtain a noise adding processing result set, wherein the noise adding processing result set comprises noise adding processing results corresponding to each image in a training image data set; then according to the noise processing result set and the standard normal distribution sampling result set, calculating a loss function value set by using a loss function, wherein the loss function value set comprises a loss function value corresponding to each image in the training image data set; and finally, carrying out average calculation on each loss function value in the loss function value set to obtain an average loss function value. In this embodiment of the present application, since the training image dataset includes a plurality of monkey pox infected skin images, a monkey pox infected skin image may be input first, a noise adding result is output after noise adding, a noise adding result corresponding to the time step t in the noise adding process is randomly selected, a sampling result corresponding to the time step t is obtained from the standard normal distribution, and then a loss function value is calculated by using a loss function. Next to this, the process is carried out, Inputting the next monkey pox infected skin image, calculating the corresponding loss function value, and the like, calculating all the loss function values corresponding to all the monkey pox infected skin images in the training image data set, and then carrying out average calculation on all the loss function values to obtain the average loss function value. For example, the training image data set includes N monkey pox infected skin images, the first monkey pox infected skin image in the training image data set is input into the trained diffusion denoising probability model, and the calculated corresponding loss function value is L 1 The method comprises the steps of carrying out a first treatment on the surface of the Then inputting a second monkey pox infected skin image in the training image data set into the trained diffusion denoising probability model, and calculating the corresponding loss function value as L 2 The method comprises the steps of carrying out a first treatment on the surface of the In this way, the Nth monkey pox infected skin image in the training image data set is input into the trained diffusion denoising probability model, and the calculated corresponding loss function value is L N . Then, the average loss function value is calculated as
Referring to fig. 7, an embodiment of the present application further provides a device for detecting an image of a skin infected with a monkey pox, which can implement the method for detecting an image of a skin infected with a monkey pox, where the device includes:
The first acquisition module is used for acquiring a skin image to be detected;
the processing module is used for carrying out noise adding processing on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding processing result;
the second acquisition module is used for acquiring sampling results from the standard normal distribution;
the calculation module is used for calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
the judging module is used for judging whether the loss function value belongs to a first threshold value interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
The specific embodiment of the device for detecting the monkey pox infected skin image is basically the same as the specific example of the method for detecting the monkey pox infected skin image, and will not be described herein.
The embodiment of the application also provides electronic equipment, which comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the monkey pox infected skin image detection method when executing the computer program. The electronic equipment can be any intelligent terminal including a tablet personal computer, a vehicle-mounted computer and the like.
Referring to fig. 8, fig. 8 illustrates a hardware structure of an electronic device according to another embodiment, the electronic device includes:
the processor 801 may be implemented by a general purpose CPU (Central Processing Unit ), a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, etc. for executing related programs to implement the technical solutions provided by the embodiments of the present application;
the Memory 802 may be implemented in the form of a Read Only Memory (ROM), a static storage device, a dynamic storage device, or a random access Memory (Random Access Memory, RAM). The memory 802 may store an operating system and other application programs, and when the technical solution provided in the embodiments of the present application is implemented by software or firmware, relevant program codes are stored in the memory 802, and the processor 801 invokes the method for detecting the monkey pox infected skin image in which the embodiments of the present application are executed;
an input/output interface 803 for implementing information input and output;
the communication interface 804 is configured to implement communication interaction between the device and other devices, and may implement communication in a wired manner (e.g., USB, network cable, etc.), or may implement communication in a wireless manner (e.g., mobile network, WIFI, bluetooth, etc.);
A bus 805 that transfers information between the various components of the device (e.g., the processor 801, the memory 802, the input/output interface 803, and the communication interface 804);
wherein the processor 801, the memory 802, the input/output interface 803, and the communication interface 804 implement communication connection between each other inside the device through a bus 805.
The embodiment of the application also provides a storage medium, which is a computer readable storage medium, and the storage medium stores a computer program, and the computer program realizes the monkey pox infected skin image detection method when being executed by a processor.
The memory, as a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory remotely located relative to the processor, the remote memory being connectable to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
According to the method and the device for detecting the monkey pox infected skin image, the electronic equipment and the storage medium, the skin image to be detected is obtained, and then the noise adding treatment is carried out on the skin image to be detected through a diffusion denoising probability model trained in advance, so that a noise adding treatment result is obtained; obtaining a sampling result from the standard normal distribution, and calculating to obtain a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result; and finally judging whether the loss function value belongs to a first threshold value interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image. The detection is carried out through the pre-trained diffusion denoising probability model, the appearance information of the monkey pox virus in human morbidity is fully considered, the detection can be completed under the condition that the image data do not need to be marked, the detection performance is better, whether the skin image to be detected is a monkey pox infected skin image or not can be detected efficiently and accurately, and the detection accuracy is high.
The embodiments described in the embodiments of the present application are for more clearly describing the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application, and as those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.
It will be appreciated by those skilled in the art that the technical solutions shown in the figures do not constitute limitations of the embodiments of the present application, and may include more or fewer steps than shown, or may combine certain steps, or different steps.
The above described apparatus embodiments are merely illustrative, wherein the units illustrated as separate components may or may not be physically separate, i.e. may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
Those of ordinary skill in the art will appreciate that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof.
The terms "first," "second," "third," "fourth," and the like in the description of the present application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is merely a logical function division, and there may be another division manner in actual implementation, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including multiple instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing a program.
Preferred embodiments of the present application are described above with reference to the accompanying drawings, and thus do not limit the scope of the claims of the embodiments of the present application. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and spirit of the embodiments of the present application shall fall within the scope of the claims of the embodiments of the present application.
Claims (10)
1. A method for detecting a monkey pox infected skin image, the method comprising:
acquiring a skin image to be detected;
carrying out noise adding treatment on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding treatment result;
obtaining a sampling result from standard normal distribution;
calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
judging whether the loss function value belongs to the first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
2. The method according to claim 1, wherein the step of calculating a loss function value from the skin image to be detected, the noise addition processing result, and the sampling result is performed by the following loss function formula:
wherein ,
where Loss represents the Loss function, ε represents the sampling result obtained from the normal distribution, ε θ Representing the diffusion denoising probability model trained in advance, x 0 Representing the skin image to be detected, beta t To represent the noise variance of the T-th step, an equal-ratio array between 0 and 1 is taken, t=1, 2,3 … T, and T is the number of diffusion steps.
3. The method of claim 1, wherein the first threshold interval is determined by:
acquiring a training image dataset comprising a plurality of monkey pox infected skin images;
carrying out noise adding treatment on each image in a training image data set through the diffusion denoising probability model trained in advance, and calculating to obtain an average loss function value;
and determining a first threshold interval according to the average loss function value and the fluctuation proportion threshold.
4. A method according to claim 3, wherein the step of denoising each image in the training image dataset by means of the diffusion denoising probability model trained in advance and calculating an average loss function value comprises:
performing standardization processing on pixel values of all images in the training image data set;
Carrying out noise adding processing on all the standardized processed images through the diffusion denoising probability model trained in advance to obtain a noise adding processing result set, wherein the noise adding processing result set comprises noise adding processing results corresponding to each image in the training image data set;
according to the noise processing result set and the standard normal distribution sampling result set, calculating a loss function value set by using the loss function, wherein the loss function value set comprises a loss function value corresponding to each image in the training image data set;
and carrying out average calculation on each loss function value in the loss function value set to obtain an average loss function value.
5. The method according to claim 1, further comprising pre-training the diffuse denoising probability model, comprising in particular:
constructing the training image dataset;
performing standardization processing on all image pixel values in the training image data set;
acquiring a monkey pox infected skin image in the training image data set for training;
according to the loss function, updating parameters of the diffusion denoising probability model by using a random gradient descent method;
Acquiring a next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image data set are trained completely, and completing one iteration;
judging whether the diffusion denoising probability model training meets the ending condition or not, outputting the diffusion denoising probability model if the ending condition is met, and continuing training if the ending condition is not met.
6. The method according to claim 5, wherein the obtaining the next monkey pox infected skin image for training according to the updated parameters of the diffusion denoising probability model until all the monkey pox infected skin images in the training image dataset are trained, and after completing one iteration, the pre-training the diffusion denoising probability model further comprises:
obtaining a verification image dataset comprising a plurality of monkey pox infected skin images and a plurality of healthy skin images;
inputting the verification image data set into the diffusion denoising probability model to test the detection accuracy of the diffusion denoising probability model;
and obtaining a test result, and controlling the training process of the diffusion denoising probability model by using an early-stop method according to the test result.
7. The method according to claim 6, wherein the step of determining whether the diffuse denoising probability model training satisfies an end condition, outputting the diffuse denoising probability model if the end condition is satisfied, and continuing training if the end condition is not satisfied, specifically comprises:
acquiring a verification error of the verification image dataset;
outputting the diffusion denoising probability model when the verification error continuously rises in the training process of the preset iteration times;
and when the verification error does not continuously rise in the training process of the preset iteration times, continuing to train the diffusion denoising probability model.
8. A monkey pox infected skin image detection apparatus, the apparatus comprising:
the first acquisition module is used for acquiring a skin image to be detected;
the processing module is used for carrying out noise adding processing on the skin image to be detected through a pre-trained diffusion denoising probability model to obtain a noise adding processing result;
the second acquisition module is used for acquiring sampling results from the standard normal distribution;
the calculation module is used for calculating a loss function value according to the skin image to be detected, the noise adding processing result and the sampling result;
The judging module is used for judging whether the loss function value belongs to the first threshold interval, if so, determining that the skin image to be detected is a monkey pox infected skin image, and if not, determining that the skin image to be detected is not a monkey pox infected skin image.
9. An electronic device comprising a memory storing a computer program and a processor implementing the detection method according to any one of claims 1 to 7 when the computer program is executed by the processor.
10. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the detection method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211101846.5A CN116029968A (en) | 2022-09-09 | 2022-09-09 | Monkey pox infection skin image detection method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211101846.5A CN116029968A (en) | 2022-09-09 | 2022-09-09 | Monkey pox infection skin image detection method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116029968A true CN116029968A (en) | 2023-04-28 |
Family
ID=86071168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211101846.5A Pending CN116029968A (en) | 2022-09-09 | 2022-09-09 | Monkey pox infection skin image detection method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116029968A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116630634A (en) * | 2023-05-29 | 2023-08-22 | 北京医准智能科技有限公司 | Image processing method, device, equipment and storage medium |
-
2022
- 2022-09-09 CN CN202211101846.5A patent/CN116029968A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116630634A (en) * | 2023-05-29 | 2023-08-22 | 北京医准智能科技有限公司 | Image processing method, device, equipment and storage medium |
| CN116630634B (en) * | 2023-05-29 | 2024-01-30 | 浙江医准智能科技有限公司 | Image processing method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111709409B (en) | Face living body detection method, device, equipment and medium | |
| WO2020207377A1 (en) | Method, device, and system for image recognition model training and image recognition | |
| EP3961484A1 (en) | Medical image segmentation method and device, electronic device and storage medium | |
| CN109858555B (en) | Image-based data processing method, device, equipment and readable storage medium | |
| US11967069B2 (en) | Pathological section image processing method and apparatus, system, and storage medium | |
| US20190065906A1 (en) | Method and apparatus for building human face recognition model, device and computer storage medium | |
| US10943154B2 (en) | Systems for modeling uncertainty in multi-modal retrieval and methods thereof | |
| WO2022188773A1 (en) | Text classification method and apparatus, device, computer-readable storage medium, and computer program product | |
| CN111767366A (en) | Question and answer resource mining method and device, computer equipment and storage medium | |
| CN116824278B (en) | Image content analysis method, device, equipment and medium | |
| CN110781413A (en) | Interest point determining method and device, storage medium and electronic equipment | |
| CN113177559B (en) | Image recognition method, system, equipment and medium combining breadth and dense convolutional neural network | |
| CN113326851B (en) | Image feature extraction method and device, electronic equipment and storage medium | |
| CN115862831B (en) | Intelligent online reservation diagnosis and treatment management system and method | |
| CN116129141A (en) | Medical data processing method, apparatus, device, medium and computer program product | |
| CN116029968A (en) | Monkey pox infection skin image detection method and device, electronic equipment and storage medium | |
| CN115713256A (en) | Medical training assessment and evaluation method and device, electronic equipment and storage medium | |
| CN117523356A (en) | Multi-view semi-supervised characterization learning method for medical image group science | |
| CN115659221A (en) | Teaching quality assessment method and device and computer readable storage medium | |
| CN116958626A (en) | Image classification model training, image classification method and device and electronic equipment | |
| CN111582404B (en) | Content classification method, device and readable storage medium | |
| CN112862840B (en) | Image segmentation method, device, equipment and medium | |
| Bühler et al. | Automated hand-raising detection in classroom videos: A view-invariant and occlusion-robust machine learning approach | |
| CN115019396A (en) | Learning state monitoring method, device, equipment and medium | |
| CN114155957A (en) | Text determination method and device, storage medium and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |