This work suggests a theoretical principle about the oscillation signal decomposition, which is based on the requirement of a pure oscillation component, in which the mean zero is extracted from the signal. Using this principle, the validity and robustness of the empirical mode decomposition (EMD) method are first proved mathematically. This work also presents a modified version of EMD by the interpolation solution, which is able to improve the frequency decomposition of the signal. The result shows that it can provide a primary theoretical basis for the development of EMD. The simulation signal verifies the effectiveness of the EMD algorithm. At the same time, compared with the existing denoising algorithm, it has achieved good results in the denoising of rolling bearing fault signals. It contributes to the development and improvement of adaptive signal processing theory in the field of fault diagnosis. It provides practical value research results for the rapid development of adaptive technology in the field of fault diagnosis.

Although the existing traditional image classification methods have been widely applied in practical problems, there are some problems in the application process, such as unsatisfactory effects, low classification accuracy, and weak adaptive ability. This method separates image feature extraction and classification into two steps for classification operation. The deep learning model has a powerful learning ability, which integrates the feature extraction and classification process into a whole to complete the image classification test, which can effectively improve the image classification accuracy. However, this method has the following problems in the application process: first, it is impossible to effectively approximate the complex functions in the deep learning model. Second, the deep learning model comes with a low classifier with low accuracy. So, this paper introduces the idea of sparse representation into the architecture of the deep learning network and comprehensively utilizes the sparse representation of well multidimensional data linear decomposition ability and the deep structural advantages of multilayer nonlinear mapping to complete the complex function approximation in the deep learning model. And a sparse representation classification method based on the optimized kernel function is proposed to replace the classifier in the deep learning model, thereby improving the image classification effect. Therefore, this paper proposes an image classification algorithm based on the stacked sparse coding depth learning model-optimized kernel function nonnegative sparse representation. The experimental results show that the proposed method not only has a higher average accuracy than other mainstream methods but also can be good adapted to various image databases. Compared with other deep learning methods, it can better solve the problems of complex function approximation and poor classifier effect, thus further improving image classification accuracy.

Due to the complexity of medical images, traditional medical image classification methods have been unable to meet the actual application needs. In recent years, the rapid development of deep learning theory has provided a technical approach for solving medical image classification. However, deep learning has the following problems in the application of medical image classification. First, it is impossible to construct a deep learning model with excellent performance according to the characteristics of medical images. Second, the current deep learning network structure and training strategies are less adaptable to medical images. Therefore, this paper first introduces the visual attention mechanism into the deep learning model so that the information can be extracted more effectively according to the problem of medical images, and the reasoning is realized at a finer granularity. It can increase the interpretability of the model. Additionally, to solve the problem of matching the deep learning network structure and training strategy to medical images, this paper will construct a novel multiscale convolutional neural network model that can automatically extract high‐level discriminative appearance features from the original image, and the loss function uses the Mahalanobis distance optimization model to obtain a better training strategy, which can improve the robust performance of the network model. The medical image classification task is completed by the above method. Based on the above ideas, this paper proposes a medical classification algorithm based on a visual attention mechanism‐multiscale convolutional neural network. The lung nodules and breast cancer images were classified by the method in this paper. The experimental results show that the accuracy of medical image classification in this paper is not only higher than that of traditional machine learning methods but also improved compared with other deep learning methods, and the method has good stability and robustness.

Road traffic signs can improve the pressure of environmental traffic, and the real-time accurate recognition of traffic signs is conducive to the promotion and development of intelligent vehicles. However, in various complex application scenarios such as tilted and deformed traffic signs, there still have the following problems: First, the feature selection of the existing traffic recognition model only considers its network layer information and fails to retain the salient feature information with strong discriminatory power effectively, and the feature enhancement is easy to introduce background noise. Second, the existing traffic sign recognition models are difficult to deploy on mobile devices and systems, resulting in weak learning capability of feature representation in deep learning backbone network, and low robustness of the high-level feature fusion performance. Therefore, we propose a cascade attention mechanism, which can associate a series of attention units using a cascade approach, and design a cascade attention feature enhancement module, which can effectively improve the feature selection and feature enhancement performance in the traffic sign recognition process. Then, we design a lightweight deep learning model and propose modal fusion, a high-level feature-guided feature refinement mechanism, and a mutual attention enhancement module. In addition, the interrelationship between traffic signs and deep modal features is given, which can effectively enhance the feature representation learning ability of the lightweight deep learning model and significantly improve the efficiency and robustness of the model. The experimental results show that our proposed method achieves the most excellent recognition results on several traffic sign datasets.