Robust radar waveform recognition algorithm based on random projections and sparse classification

The recently introduced theory of compressive sensing enables the recovery of sparse or compressible signals from a small set of nonadaptive, linear measurements. If properly chosen, the number of measurements can be much smaller than the number of Nyquist-rate samples. Interestingly, it has been shown that random projections are a near-optimal measurement scheme. This has inspired the design of hardware systems that directly implement random measurement protocols. However, despite the intense focus of the community on signal recovery, many (if not most) signal processing problems do not require full signal recovery. In this paper, we take some first steps in the direction of solving inference problems-such as detection, classification, or estimation-and filtering problems using only compressive measurements and without ever reconstructing the signals involved. We provide theoretical bounds along with experimental results.

For passive radar detection system, radar waveform recognition is an important research area. In this paper, we explore an automatic radar waveform recognition system to detect, track and locate the low probability of intercept (LPI) radars. The system can classify (but not identify) 12 kinds of signals, including binary phase shift keying (BPSK) (barker codes modulated), linear frequency modulation (LFM), Costas codes, Frank code, P1-P4 codesand T1-T4 codeswith a low signal-to-noise ratio (SNR). It is one of the most extensive classification systems in the open articles. A hybrid classifier is proposed, which includes two relatively independent subsidiary networks, convolutional neural network (CNN) and Elman neural network (ENN). We determine the parameters of the architecture to make networks more effectively. Specifically, we focus on how the networks are designed, what the best set of features for classification is and what the best classified strategy is. Especially, we propose several key features for the classifier based on Choi–Williams time-frequency distribution (CWD). Finally, the recognition system is simulated by experimental data. The experiments show the overall successful recognition ratio of 94.5% at an SNR of −2 dB.

Automotive radars, with a widespread emergence in the last decade, have faced various jamming attacks. Utilizing low probability of intercept (LPI) radar waveforms, as one of the essential solutions, demands an accurate waveform recognizer at the intercept receiver. Numerous conventional approaches have been studied for LPI radar waveform recognition, but their performance is inadequate under channel condition deterioration. In this letter, by exploiting deep learning (DL) to capture intrinsic radio characteristics, we propose a convolutional neural network (CNN), namely LPI-Net, for automatic radar waveform recognition. In particular, radar signals are first analyzed by a time-frequency analysis using the Choi-Williams distribution. Subsequently, LPI-Net, primarily consisting of three sophisticated modules, is built to learn the representational features of time-frequency images, in which each module is constructed with a preceding maps collection to gain feature diversity and a skip-connection to maintain informative identity. Simulation results show that LPI-Net achieves the 13-waveform recognition accuracy of over 98% at 0 dB SNR and further performs better than other deep models.

Compressive sensing theory asserts that, under certain conditions, a high dimensional but compressible signal can be recovered from a small number of random linear projections by utilizing computationally efficient algorithms. The a priori knowledge of the basis in which the signal of interest is sparse is the key assumption utilized by such algorithms. However, the basis in which the signal is the sparsest is unknown for many natural signals of interest. Instead there may exist multiple bases which lead to a compressible representation of the signal: e.g., an image is compressible in different wavelet transforms. We show that a significant performance improvement can be achieved by utilizing multiple estimates of the signal using sparsifying bases in the context of signal reconstruction from compressive samples. Further, we derive a customized interior-point method to jointly obtain multiple estimates of a 2-D signal (image) from compressive measurements utilizing multiple sparsifying bases as well as the fact that the images usually have a sparse gradient.

In the compressive sampling theory, a small number of random linear projections of a sparse or compressible signal have contained sufficient information and the original signal can be accurately reconstructed by taking advantage of modern optimization algorithms. We proposed an approximate l 0 norm based signal reconstruction algorithm in this paper. It not only can convert the classical constrained l 0 minimization problem of the compressive sampling theory into an unconstrained optimization problem, but also can reduce the dimension of the search space substantially. The experiment results have shown that our proposed algorithm can improve the sparse signal reconstruction performance while maintaining appropriate signal reconstruction efficiency.

One major challenge of QRS wave detection is the deleterious impact of inevitable noise in electrocardiogram (ECG) signals, which can deteriorate the detection performance of algorithms. Previous research has shown a nonlinear monostable potential‐based algorithm for enhancing the QRS complex in a noisy environment. Its main mechanism is the stochastic resonance effect, which transfers noise energy to information‐carrying ECG signals in the nonlinear potential. However, the QRS detection algorithm based on monostable stochastic resonance (MSR) exhibits a limited noise margin to maintain extremely good performance. To further improve noise robustness, we propose a multi‐parameter constrained bistable SR (MCBSR) module for the QRS detection algorithm without requiring a large amount of training data. The MCBSR module can maximize the ECG signal while suppressing noise when there is a QRS complex and minimizes the ECG signal otherwise. Using four publicly available databases for testing, the MCBSR‐based QRS detection algorithm has been proven to achieve superior performance among state‐of‐the‐art algorithms. Compared to the conventional nonlinear method, the MCBSR enables the QRS detection algorithm to have significantly stronger noise robustness under various noise types. Therefore, this study successfully provides a more excellent nonlinear potential and greatly enhances the noise robustness of the QRS detection algorithm in large noise environments. © 2025 The Author(s). IEEJ Transactions on Electrical and Electronic Engineering published by Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.

An universal method for wordform recognition and generation is proposed. It includes the model of word formation in natural languages and algorithms for wordform generation and recognition. In the wordformation model, abstracting from typical notions in the morphology of natural languages takes place Actual problems of machine translation, spelling checking, text analysis and knowledge extraction from them, computer–user dialog in a natural language, analysis of queries in information search systems of global information networks include multilevel automatic text processing. This levels are morphological, syntax and semantic. There are a large number of methods for solving the morphological level problems in text processing (see short review in [1]). However, they are designed for one or several languages that are similar in their word formation. At the same time, in connection with worldwide globalization, it is topical to develop applications for text pro cessing in different languages. Therefore, the necessity arises to create a method for morphological processing of texts in natural languages of different groups and families. This work proposes a universal method for the generation and recognition of word forms independent of the formation type and capable of processing the entire paradigm of a word. The method includes a model for the formation and wordform generation and recognition algorithm. The Model of formation in a natural language (below the model of formation) implies that the generation of any word form with a given grammatical meaning can be presented as the sequence of a finite number of transformations of the base. The method of presenting formation, as well the generation and recognition algorithms are interrlated. On the one hand, one part governs the other. On the other, simplification of one part causes complication of the other. The objective of the word form generation algorithm is to obtainthe word form F corresponding to the base S and grammatical meaning G. The result of word form generation is always unambiguous. The input data are S and G. The output data are F. The generation algorithm consists of two stages. At the first stage a required transformation chain R is searched using the input data, the base S, and grammatical meaning G. The second stage is the application of the found chain R to the base S to obtain the required form F. The wordform recognition algorithm recognizes the grammatical meaning G and the base S that correspond to the starting word form F. The input data are F. The output data are S and G. This algorithm consists in searching the chains of transformations in a cycle. At each iteration, a new chain Ri is taken as well as a respective base type Ti and a grammatical meaning Gi. Further, the next reverse chain is applied to the starting word form F to obtain the base Si. If the same base type Ti corresponds to the found base Si and the next chain then the result that consists of the obtained chain Si and the grammatical meaning Giis added to the obtained results. Thus, the iteration is completed and a transition to the next iteration occurs. The number of iterations equals n, the number of the chains of transformations. Thus, we have developed a universal method for wordform generation and recognition that includes the model of formation and the processing algorithm. The proposed processing algorithms are applicable to natural languages of different groups and families. Algorithms for the Russian, German, Spanish, and Finnish languages were successfully tested. The transformation chains were classified and the algorithm of automated construction of transformation chains was proposed based on this classification.

In this paper, we deal with a face recognition method for the emotional face images. Since the face recognition is one of the most natural and straightforward biometric methods, there have been various research works. However, most of them are focused on the expressionless face images and have had a very difficult problem if we consider the facial expression. In real situations, however, it is required to consider the emotional face images. Here, three basic human emotions such as happiness, sadness, and anger are investigated for the face recognition. And, this situation requires a robust face recognition algorithm then we use a fuzzy Fisher’s Linear Discriminant (FLD) algorithm with the wavelet transform. The fuzzy Fisherface is a statistical method that maximizes the ratio of between-scatter matrix and within-scatter matrix and also handles the fuzzy class information. The experimental results obtained for the CBNU face databases reveal that the approach presented in this paper yields better recognition performance in comparison with the results obtained by other recognition methods.

In order to improve the recognition ability of tunnel cracks in the UAV platform with a vision imaging system in the UAV platform with a vision imaging system, this paper proposes a tunnel crack segmentation algorithm using SPGD-and-generative adversarial network fusion. The SPGD algorithm can enhance the detail and edge information of a tunnel crack image, which improves the clarity of the tunnel crack image. The new generative adversarial network (GAN) is designed by using an improved U-Net generator and full convolutional network (FCN) discriminator to form a new network; the improved generative adversarial network can effectively segment tunnel crack images after stochastic parallel gradient descent (SPGD) algorithm processing, especially the texture feature extraction and segmentation of small tunnel cracks, which can improve the rate of recognition of tunnel cracks. Based on collected tunnel crack image data, we selected 12 typical tunnel crack images and verified the rationality and advanced nature of the proposed recognition algorithm by comparing it with other recognition methods. The results show that the recognition rate of the proposed tunnel crack recognition algorithm was significantly improved.

A robust weld seam recognition method under heavy noise based on structured-light vision

Adding spatial penalty to fuzzy C-means (FCM) model is an important way to reduce the influence of noise in image segmentation. However, these improved algorithms easily cause segmentation failures when the image has the characteristics of unequal cluster sizes. Besides, they often fall into local optimal solutions if the initial cluster centers are improper. This paper presents a noise robust hybrid algorithm for segmenting image with unequal cluster sizes based on chaotic crow search algorithm and improved fuzzy c-means to overcome the above defects. Firstly, each size of clusters is integrated into the objective function of noise detecting fuzzy c-means algorithm (NDFCM), which can reduces the contribution of larger clusters to objective function and then the new membership degree and cluster centers are deduced. Secondly, a new expression called compactness, representing the pixel distribution of each cluster, is introduced into the iteration process of clustering. Thirdly, we use two- paths to seek the optimal solutions in each step of iteration: one path is produced by the chaotic crow search algorithm and the other is originated by gradient method. Furthermore, the better solutions of the two-paths go to next generation until the end of the iteration. Finally, the experiments on the synthetic and non–destructive testing (NDT) images show that the proposed algorithm behaves well in noise robustness and segmentation performance.

In order to realize high-accuracy recognition of aerobics actions, a highly applicable deep learning model and faster data processing methods are required. Therefore, it is a major difficulty in the field of research on aerobics action recognition. Based on this, this paper studies the application of the convolution neural network (CNN) model combined with the pyramid algorithm in aerobics action recognition. Firstly, the basic architecture of the convolution neural network model based on the pyramid algorithm is proposed. Combined with the application strategy of the common recognition model in aerobics action recognition, the traditional aerobics action capture information is processed. Through the characteristics of different aerobics actions, different accurate recognition is realized, and then, the error of the recognition model is evaluated. Secondly, the composite recognition function of the convolution neural network model in this application is constructed, and the common data layer effect recognition method is used in the optimization recognition. Aiming at the shortcomings of the composite recognition function, the pyramid algorithm is used to improve the convolution neural network recognition model by deep learning optimization. Finally, through the effectiveness comparison experiment, the results show that the convolution neural network model based on the pyramid algorithm is more efficient than the conventional recognition method in aerobics action recognition.

An algorithm for the robust detection and recognition of gestures for the interaction between human and a domestic floor cleaner robot is presented. The gestures are selected through a user study, in which the participants are asked to show natural gestures to the robot in given specific interaction scenarios. The gestures selected are those repeated by majority of participants and consist both commanding (say start cleaning) and social interaction (say greeting) gestures. The gesture recognition algorithm is developed using a combination of robust angular, positional, and directional features. The frontal and sagittal plains of human body are identified and invariant angular features are extracted from the skeletal data of a Kinect sensor. Robust positional and directional features are extracted by skeletal reconstruction using the invariant angular features and link rotation matrices. Dynamic time warping of features is done to make the algorithm robust to gesturing speed. Gestures are detected and recognized by calculating multiclass probability estimates by pairwise coupling method. The algorithm provided 97.26% recognition accuracy for a ten class robot commanding gesture database collected from multiple subjects.

An effective way to increase noise robustness in automatic speech recognition (ASR) systems is feature enhancement based on an analytical distortion model that describes the effects of noise on the speech features. One of such distortion models that has been reported to achieve a good trade-off between accuracy and simplicity is the masking model. Under this model, speech distortion caused by environmental noise is seen as a spectral mask and, as a result, noisy speech features can be either reliable (speech is not masked by noise) or unreliable (speech is masked). In this paper, we present a detailed overview of this model and its applications to noise robust ASR. Firstly, using the masking model, we derive a spectral reconstruction technique aimed at enhancing the noisy speech features. Two problems must be solved in order to perform spectral reconstruction using the masking model: (1) mask estimation, i.e. determining the reliability of the noisy features, and (2) feature imputation, i.e. estimating speech for the unreliable features. Unlike missing data imputation techniques where the two problems are considered as independent, our technique jointly addresses them by exploiting a priori knowledge of the speech and noise sources in the form of a statistical model. Secondly, we propose an algorithm for estimating the noise model required by the feature enhancement technique. The proposed algorithm fits a Gaussian mixture model to the noise by iteratively maximising the likelihood of the noisy speech signal so that noise can be estimated even during speech-dominating frames. A comprehensive set of experiments carried out on the Aurora-2 and Aurora-4 databases shows that the proposed method achieves significant improvements over the baseline system and other similar missing data imputation techniques.

This paper introduces the extraction of speech features realizing noise robustness for speech recognition. It also explores advanced speech analysis techniques named RSF (Running Spectrum Filtering)/DRA (Dynamic Range Adjustment) in detail. The new experiments on phase recognition were carried out using 40 male and female speakers for training and 5 other male and female speakers for recognition. The result of recognition rate is improved from 17% to 63% under car noise at -10dB SNR for example. It shows the high noise robustness of the proposed system. In addition, the new parallel/pipelined LSI design of the system is proposed. It considerably reduces the calculation time. Using this architecture, the real time speech recognition can be developed. For this system, both of full-custom LSI design and FPGA design are introduced.