Image classification method and system based on similarity uncertainty optimization

Abstract

The present invention discloses an image classification method and system based on similarity uncertainty optimization, belonging to the field of computer vision technology. The method comprises: obtaining an image to be processed; processing the image to be processed using a trained image classification model to obtain an image classification result; and, when training the image classification model, quantifying the similarity uncertainty between difficult samples in a noisy training set and dynamic class prototypes, and performing backpropagation optimization on the constructed network in combination with the classification loss and the divergence evaluation loss calculated from clean samples in the training set. This method can learn more discriminative information from samples with higher uncertainty, improving the robustness of the model and resolving the problem of noise data affecting the prediction stability of existing models during training.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to an image classification method and system based on similarity uncertainty optimization. Background Art

[0002] The statements in this section merely mention background art related to the present invention and do not necessarily constitute prior art.

[0003] Noisy label learning is an important research direction in machine learning, especially in supervised learning tasks. Traditional supervised learning relies on high-quality labeled datasets, but in practice, obtaining completely accurate labeled data is often very difficult, especially for large-scale datasets.

[0004] Due to human error, sensor failure, or limitations of automatic labeling systems, data labels may contain noise, meaning that the labels do not match the true categories. This noise not only affects the generalization ability of the model but may also cause the model to overfit to the incorrect labels, thereby reducing its performance.

[0005] The development of noisy label learning can be divided into several stages. Early research focused on mitigating the impact of noise on model training by modeling it. This typically assumed that noise was generated from a certain distribution and used statistical or Bayesian methods to construct a noise transfer matrix to infer the true label. Subsequently, with the rise of deep learning, noisy label learning methods gradually shifted to more complex models and techniques, such as those based on loss function modification, data cleaning, and model robustness enhancement. Some methods reduce the impact of noise on gradients by adjusting the loss function, while others improve data quality by cleaning the data and removing or correcting incorrect labels.

[0006] In recent years, dual-model training, semi-supervised learning, and self-supervised learning techniques have made significant progress in learning with noisy labels. These methods leverage unlabeled data or leverage cross-learning between models to mitigate the effects of noise. However, none of these methods consider the potential for inherent noise in the data, which often leads to overfitting or performance degradation, compromising model accuracy and robustness.

[0007] Furthermore, existing methods attempt to address the aforementioned issues by selecting or reweighting samples, but they ignore the effects of perceptual ambiguity and bias caused by the inherent noise contained in the dataset and rely heavily on softmax classifiers, neglecting the exploration of knowledge in low-quality samples. Summary of the Invention

[0008] In order to address the deficiencies of the prior art, the present invention provides an image classification method, system, electronic device, computer-readable storage medium and computer program product based on similarity uncertainty optimization, which utilizes the inherent noise in uncertain quantization data sets to improve the accuracy and robustness of image classification.

[0009] In a first aspect, the present invention provides an image classification method based on similarity uncertainty optimization;

[0010] An image classification method based on similarity uncertainty optimization, comprising:

[0011] Get the image to be processed;

[0012] Processing the image to be processed using a trained image classification model to obtain an image classification result;

[0013] When training the image classification model, the uncertainty of the similarity between difficult samples in the noisy training set and the dynamic class prototype is quantified, and the constructed network is back-propagated and optimized in combination with the classification loss and the divergence evaluation loss calculated by the clean samples in the training set.

[0014] In some embodiments, quantifying the uncertainty of the similarity between the difficult samples in the noisy training set and the dynamic class prototype specifically includes:

[0015] Based on the difficult sample and the dynamic class prototype, a similarity uncertainty distribution is constructed by combining cosine similarity calculation and graph convolutional neural network;

[0016] Use the Monte Carlo algorithm to sample from the similarity uncertainty distribution and calculate the similarity uncertainty loss;

[0017] The dynamic class prototype is constructed using clean samples.

[0018] In some embodiments, the divergence evaluation loss when training the image classification model is calculated using clean samples in the training set as follows: the clean samples are processed through the MLP network to obtain the clean score corresponding to each of the clean samples, and the divergence evaluation loss is calculated using the dual network prediction divergence as supervision.

[0019] In some embodiments, calculating the classification loss when training the image classification model is specifically as follows: performing dynamic label masking according to the confidence of each sample and calculating the classification loss;

[0020] The labels corresponding to the difficult samples are pseudo labels corrected by linear addition of cosine similarity and dual network prediction.

[0021] In some embodiments, training the image classification model further includes: extracting features from the training set, inputting the extracted features into the GMM algorithm for processing, obtaining the confidence corresponding to each sample, and dividing the samples based on the confidence.

[0022] In some embodiments, back propagation optimization of the constructed network is specifically performed by linearly adding the classification loss, similarity uncertainty loss, and divergence evaluation loss, and calculating the gradient to perform back propagation optimization of the network.

[0023] In a second aspect, the present invention provides an image classification system based on similarity uncertainty optimization;

[0024] An image classification system based on similarity uncertainty optimization, comprising:

[0025] The acquisition module is configured to: acquire an image to be processed;

[0026] The image classification module is configured to: process the image to be processed using a trained image classification model to obtain an image classification result;

[0027] When training the image classification model, the uncertainty of the similarity between difficult samples in the noisy training set and the dynamic class prototype is quantified, and the constructed network is back-propagated and optimized in combination with the classification loss and the divergence evaluation loss calculated by the clean samples in the training set.

[0028] In a third aspect, the present invention provides an electronic device;

[0029] An electronic device includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above-mentioned image classification method based on similarity uncertainty optimization.

[0030] In a fourth aspect, the present invention provides a computer-readable storage medium;

[0031] A computer-readable storage medium stores a computer program / instruction thereon, which, when executed by a processor, implements the steps of the above-mentioned image classification method based on similarity uncertainty optimization.

[0032] In a fifth aspect, the present invention provides a computer program product;

[0033] A computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps of the above-mentioned image classification method based on similarity uncertainty optimization.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. The technical solution provided by the present invention quantifies the similarity uncertainty of fixed noise in the data set, and quantifies the uncertainty in the sample by modeling the similarity distribution between difficult samples and class prototypes through cosine similarity and graph convolutional neural network, thereby helping the model to better utilize the information in low-quality samples.

[0036] 2. To address the network's overconfidence problem, the technical solution provided by this invention implements a dynamic masking strategy for each sample's prediction and calculates the classification loss for the masked prediction labels. This combined loss optimizes the network, improving its classification performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0038] Figure 1 A schematic diagram of a process flow provided by an embodiment of the present invention;

[0039] Figure 2 A detailed flowchart provided for an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of a process for calculating similarity uncertainty loss provided by an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of a prediction process for divergence assessment loss provided in an embodiment of the present invention;

[0042] Figure 5 An example diagram of image classification provided by an embodiment of the present invention;

[0043] Figure 6 A schematic diagram of the process of performing sample partitioning using the GMM algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0046] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0047] Explanation of terms:

[0048] Noisy label learning: Noisy label learning is a technique used in machine learning to address incorrect or inaccurate labeling in datasets. Because real-world datasets may contain noisy labels, which can affect model performance, noisy label learning uses methods such as algorithm tuning, data cleaning, and model weighting to mitigate the negative impact of noisy labels on model training and improve the model's generalization and robustness.

[0049] Uncertainty generally consists of two types: structural uncertainty, which refers to poor model performance due to issues like the chosen model structure and parameters. This uncertainty can often be addressed by increasing the sample size. Epistemic uncertainty, on the other hand, arises from inherent noise in the dataset.

[0050] Graph Convolutional Networks (GCNNs): A model specifically designed to process graph-structured data. It updates node representations by aggregating node features with information about neighboring nodes. It effectively captures the relationships between nodes and edges in a graph and is widely used in fields such as social network analysis, recommender systems, and bioinformatics.

[0051] ResNet18: A deep convolutional neural network based on residual networks with a depth of 18 layers. It alleviates the gradient vanishing problem in deep networks by introducing skip connections, making training deeper networks more efficient.

[0052] A class prototype is a central point or feature vector representing a class or cluster, describing the typical characteristics or patterns of that class. Common class prototypes include centroids or other features that characterize the distribution of data within a cluster, helping to measure the similarity of new data points to the class.

[0053] Similarity: A metric used to measure the degree of similarity between two objects, usually expressed as a numerical value. It is used to compare the characteristics of objects; higher values ​​indicate more similar objects. Common calculation methods include Euclidean distance and cosine similarity.

[0054] GMM: A probabilistic model that models data as a mixture of multiple Gaussian distributions, commonly used in soft clustering. Using the expectation-maximization (EM) algorithm, GMM estimates the parameters of each Gaussian distribution and is widely used in tasks such as clustering and anomaly detection.

[0055] Clean samples: samples that are considered to be correctly labeled.

[0056] Difficult samples: Samples that are difficult to classify and whose labels are often wrong.

[0057] MLP (Multi-layer Perceptron): It is a feedforward neural network consisting of an input layer, one or more hidden layers, and an output layer. The neurons in each layer are connected by weights and nonlinear activation functions. It is widely used in classification and regression tasks.

[0058] Example 1

[0059] The uncertainty of noise labels in existing image classification models leads to overfitting of the model, affecting the accuracy and robustness of image classification; therefore, the present invention provides an image classification method based on similarity uncertainty optimization.

[0060] Next, combine Figures 1-6 , a method for image classification based on similarity uncertainty optimization disclosed in this embodiment is described in detail. The method for image classification based on similarity uncertainty optimization includes the following steps:

[0061] S1. Obtain the image to be processed.

[0062] In this embodiment, the images to be processed are RGB images with a size of 32*32, and each image contains an object to be classified.

[0063] S2. Process the image to be processed through the trained image classification model to obtain the image classification result.

[0064] In this embodiment, the image classification model is a ResNet18 model, the last layer of which is a fully connected layer capable of classifying images; the input of the ResNet18 model is the image to be processed, and the output is the image classification result.

[0065] Image classification results refer to the model outputting a prediction score for each image, corresponding to each category. The prediction score represents the probability that the image belongs to the corresponding category. The category with the highest probability is selected as the model's prediction result, indicating that the image belongs to that category. The presence of label noise increases data uncertainty, making it difficult for the model to distinguish between accurate and incorrect sample labels. Often, noisy labels can make the model's predictions for certain samples more unstable because the model may mistakenly learn from noisy labels as valid signals. This uncertainty stems from multiple factors, including noise in the data itself, inaccurate labeling, and the model's tendency to overfit on noisy samples.

[0066] This makes uncertainty quantification and modeling the key to solving the noisy label problem. Therefore, in this embodiment, during the model training process, this part of uncertainty is quantified and utilized to solve the noisy label problem.

[0067] Therefore, in this embodiment, two initial image classification models are constructed, and the architecture and parameters of the initial image classification models are the same. Then, the two initial image classification models are trained in parallel based on the noise label learning method with similarity uncertainty optimization, and the training method is the same. Finally, the trained models are evaluated, and the initial image classification model with higher classification accuracy is used as the final image classification model.

[0068] As an implementation method, the specific process of training the image classification model is as follows:

[0069] Step 1: Build two parallel neural networks.

[0070] Step 2: Obtain a noisy data set as a training set, and extract features from the training set using the feature extractor in the initial image classification model.

[0071] In this embodiment, the feature extractor is a ResNet18 network.

[0072] Step 3: The extracted features are fed into the GMM algorithm for processing. The GMM algorithm calculates a confidence score for each sample. When the confidence score exceeds a given threshold, these samples are classified as clean samples, and the remaining samples are classified as hard samples. For the clean sample set, steps 4 and 5 are performed in parallel; for the hard sample set, steps 6 and 7 are performed in parallel.

[0073] Combine Figure 6 Specifically, the extracted features are sample normalized to obtain the initialized mean vector, covariance matrix and mixing coefficient, and then it is determined whether the preset number of iterations is reached; if so, the confidence score of each sample is output; if not, the EM algorithm is used for iterative calculation.

[0074] In this embodiment, no improvement is made to the existing GMM algorithm, and the process is automatically completed by calling the existing algorithm, which will not be described in detail here.

[0075] Step 4: Input the clean sample set into the MLP network. The MLP network calculates a clean score for each clean sample, which is the probability that the sample is a clean sample. The predictions of the dual networks are used to calculate the JS divergence as the probability that the sample is approximately clean, and this probability is used as supervision to calculate the prediction divergence evaluation loss.

[0076] In this example, an MLP network is used to process the clean sample set to assess the importance of different clean samples. However, due to the inherent noise in the data, the true probability of the sample is unknown, making MLP supervision difficult. To address this issue, this example proposes using the Jensen-Shannon (JS) divergence to approximate the true noise type of each clean sample and train the MLP accordingly.

[0077] The training process involves two networks training in parallel. By using the JS divergence to estimate the prediction inconsistency between the two networks as an approximation of the true probability of clean samples, the prediction divergence can be optimized more smoothly. The JS divergence calculation formula is expressed as follows:

[0078]

[0079] Where P(i) and Q(i) represent the prediction results of two parallel networks respectively.

[0080] Step 5: Perform a certain degree of dynamic label masking on the prediction of each sample in the clean sample set based on the confidence level and calculate the classification loss. The classification loss function of the clean sample set is as follows:

[0081]

[0082] Among them, L ce is the cross entropy loss commonly used in the field of noise labels, Indicates a label, Represents the model's predictions.

[0083] Models commonly suffer from overconfidence, a phenomenon in which models display unreasonably high confidence in their incorrect predictions. Specifically, a model may give very high prediction probabilities for certain input samples, even when these predictions are incorrect. Therefore, in this embodiment, the model's predictions are masked to a certain extent based on confidence levels to combat this overconfidence.

[0084] Step 6: Calculate the class prototype based on the clean sample, quantify the similarity uncertainty between the difficult sample and the dynamic class prototype, and calculate the similarity uncertainty loss.

[0085] In label noise learning (LNL), noisy labels often originate from low-quality samples. Existing methods often ignore the differences between these samples and class prototypes, affecting overall performance. In this embodiment, samples in difficult datasets are treated as low-quality data. By characterizing the uncertainty of the similarity between difficult samples and class prototypes, perceptual ambiguity and bias are suppressed. Specifically, the similarity representation between difficult samples and prototypes is converted from a deterministic scalar to a probabilistic representation, naturally incorporating this uncertainty into the loss calculation for better model optimization.

[0086] For example, the specific process is as follows:

[0087] (1) Calculate the class prototype based on the clean sample and calculate the cosine similarity between the difficult sample and the class prototype as the mean of their similarities.

[0088] Specifically, the class prototype calculation based on the clean sample is as follows: the confidence obtained in step 3 is used as the weight coefficient, and the class prototype calculated in the previous training stage is dynamically weighted to obtain the class prototype. It can be expressed as:

[0089]

[0090] Where, is the class prototype of the k-th category in the b-th training stage, is the number of elements belonging to the kth class in the clean set, w i is the confidence of the corresponding sample, f φ (x i ) are the characteristics of the corresponding samples, is the class prototype of the k-th category in the b-1-th training stage.

[0091] (2) Use graph convolutional neural networks to calculate the variance between difficult samples and each class, and use the mean and variance to construct the similarity uncertainty distribution.

[0092] Specifically, the mean of the similarity between the difficult sample and the class prototype in step (1) is input as a node into the graph convolutional neural network, processed in sequence by multiple convolutions, batch normalization and activation functions, and then passed through the softplus layer to ensure that the output is positive. The output result is normalized so that the output result of GCN is the variance of the difficult sample corresponding to each category.

[0093] A normal distribution is constructed based on the mean and variance of each type corresponding to the difficult samples obtained above, and the normal distribution is used as the similarity uncertainty distribution.

[0094] Normal distribution is a distribution that satisfies N(μ,σ 2), where μ is the mean, σ 2 is the variance.

[0095] (3) Use the Monte Carlo algorithm to sample from the similarity uncertainty distribution and calculate the similarity uncertainty loss.

[0096] In this embodiment, the mean and variance of each type corresponding to the obtained difficult samples are used as the similarity and uncertainty of the samples, respectively. However, the gradient calculated directly from the distribution constructed by the mean and variance is 0, which is difficult to use for network optimization.

[0097] Therefore, in this step, the mean and variance of each type corresponding to the obtained difficult samples are reparameterized to obtain a quantized uncertainty value, that is, a differentiable normal distribution; then, the differentiable normal distribution is sampled by the Monte Carlo algorithm to calculate the uncertainty loss; the above process is repeated multiple times, and the mean of all the uncertainty losses obtained is used as the final uncertainty loss.

[0098] The uncertainty value is expressed as:

[0099] s=μ+λ*σ 2 ;

[0100] Where s represents the uncertainty value, μ represents the calculated mean, and σ 2 represents the calculated variance, and λ represents a random variable from the standard normal distribution.

[0101] The calculation formula of uncertainty loss is as follows:

[0102]

[0103] In the formula, x represents the sample, k is the category to which the sample belongs, and s k,t and s j,t It represents the uncertainty value obtained by sampling through the Monte Carlo algorithm, and N is the total number of categories.

[0104] Step 7: Use the linear sum of cosine similarity and dual-model predictions as the pseudo-label for difficult samples, perform dynamic label masking, and calculate the classification loss.

[0105] Among them, the classification loss function of difficult samples is expressed as follows:

[0106]

[0107] Where, represents the obtained pseudo label, Represents the model's predictions.

[0108] Based on this, cosine similarity is used to correct pseudo labels for difficult samples to obtain more realistic pseudo labels.

[0109] Step 8: Linearly sum all losses and calculate the gradient to perform backpropagation optimization on the network.

[0110] Step 9: Determine whether the set number of cycles has been reached. If it has, the network at this time is the trained initial image classification model; otherwise, continue the optimization cycle.

[0111] Step 10: The one with higher accuracy among the two trained initial image classification models is used as the final image classification model.

[0112] In summary, this embodiment first uses the GMM method to classify samples into clean samples and hard samples. Then, a dynamic class prototype is constructed from the clean samples, and uncertainty is quantified based on the similarity between the class prototype and the clean samples, thereby suppressing perceptual ambiguity and bias in noisy data. Prediction divergence evaluation is used to assess the importance of clean samples, imposing stricter constraints on the model. The model is optimized using an uncertainty perception loss function, a masked classification loss function, and a prediction divergence evaluation loss function, making it more robust to noisy data.

[0113] Example 2

[0114] This embodiment discloses an image classification system based on similarity uncertainty optimization, including:

[0115] The acquisition module is configured to: acquire an image to be processed;

[0116] The image classification module is configured to: process the image to be processed using a trained image classification model to obtain an image classification result;

[0117] When training the image classification model, the uncertainty of the similarity between difficult samples in the noisy training set and the dynamic class prototype is quantified, and the constructed network is back-propagated and optimized in combination with the classification loss and the divergence evaluation loss calculated by the clean samples in the training set.

[0118] It should be noted that the acquisition module and image classification module described above correspond to the steps in Example 1. The examples and application scenarios implemented by the modules and the corresponding steps are the same, but are not limited to the contents disclosed in Example 1. It should be noted that the modules described above, as part of a system, can be executed in a computer system, such as a set of computer-executable instructions.

[0119] Example 3

[0120] A third embodiment of the present invention provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the above-mentioned image classification method based on similarity uncertainty optimization are completed.

[0121] Example 4

[0122] A fourth embodiment of the present invention provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the above-mentioned image classification method based on similarity uncertainty optimization are completed.

[0123] Example 5

[0124] A fifth embodiment of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the above-mentioned image classification method based on similarity uncertainty optimization.

[0125] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide the functions for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0128] The descriptions of the various embodiments in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0129] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. An image classification method based on similarity uncertainty optimization, characterized in that: include: Get the image to be processed; Processing the image to be processed using a trained image classification model to obtain an image classification result; When training the image classification model, the uncertainty of the similarity between difficult samples in the noisy training set and dynamic class prototypes is quantified, and the constructed network is optimized by backpropagation combined with the classification loss and the divergence evaluation loss calculated using the clean samples in the training set; The uncertainty of the similarity between the difficult samples in the noisy training set and the dynamic class prototype specifically includes: Based on the difficult sample and the dynamic class prototype, a similarity uncertainty distribution is constructed by combining cosine similarity calculation and graph convolutional neural network; Use the Monte Carlo algorithm to sample from the similarity uncertainty distribution and calculate the similarity uncertainty loss; The dynamic class prototype is constructed using clean samples; The divergence evaluation loss when training the image classification model by calculating the clean samples in the training set is specifically as follows: the clean samples are processed by the MLP network to obtain the clean score corresponding to each clean sample, and the divergence evaluation loss is calculated using the dual network prediction divergence as supervision.

2. The image classification method based on similarity uncertainty optimization according to claim 1, characterized in that: Calculating the classification loss when training the image classification model is specifically as follows: performing dynamic label masking according to the confidence of each sample and calculating the classification loss; The labels corresponding to the difficult samples are pseudo labels corrected by linear addition of cosine similarity and dual network prediction.

3. The image classification method based on similarity uncertainty optimization according to claim 1, characterized in that: Training the image classification model also includes: extracting features from the training set, inputting the extracted features into the GMM algorithm for processing, obtaining the confidence level corresponding to each sample, and dividing the samples based on the confidence level.

4. The image classification method based on similarity uncertainty optimization according to claim 1, characterized in that: The back propagation optimization of the constructed network is specifically as follows: linearly add the classification loss, similarity uncertainty loss and divergence evaluation loss, and calculate the gradient to perform back propagation optimization of the network.

5. Image classification system based on similarity uncertainty optimization, characterized in that include: The acquisition module is configured to: acquire an image to be processed; The image classification module is configured to: process the image to be processed using a trained image classification model to obtain an image classification result; When training the image classification model, the uncertainty of the similarity between difficult samples in the noisy training set and dynamic class prototypes is quantified, and the constructed network is optimized by backpropagation combined with the classification loss and the divergence evaluation loss calculated using the clean samples in the training set; The uncertainty of the similarity between the difficult samples in the noisy training set and the dynamic class prototype specifically includes: Based on the difficult sample and the dynamic class prototype, a similarity uncertainty distribution is constructed by combining cosine similarity calculation and graph convolutional neural network; Use the Monte Carlo algorithm to sample from the similarity uncertainty distribution and calculate the similarity uncertainty loss; The dynamic class prototype is constructed using clean samples; The divergence evaluation loss when training the image classification model by calculating the clean samples in the training set is specifically as follows: the clean samples are processed by the MLP network to obtain the clean score corresponding to each clean sample, and the divergence evaluation loss is calculated using the dual network prediction divergence as supervision.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the image classification method based on similarity uncertainty optimization according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the steps of the image classification method based on similarity uncertainty optimization according to any one of claims 1 to 4 are implemented.

8. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the image classification method based on similarity uncertainty optimization according to any one of claims 1 to 4 are implemented.

Images