Artificial Neural Networks For Recognition Research Articles

An Artificial Olfactory Chemical‐Resistant Synapse for Training‐Free Gas Recognition

AbstractTransferring the concept of chemical‐driven responses into artificial intelligence technology holds the key to mimicking olfactory for neuromorphic computing of chemical recognition. Currently, artificial olfactory systems are designed based on chemical sensor arrays. Time‐dependent responses of the sensor arrays are processed by artificial neural networks for recognition. However, the sensors generate instantly volatile responses, and algorithms for the processing of the time‐dependent responses have not been involved. The recognition accuracy and speed are severely impeded. A sensor array can only achieve an accuracy of 90% after at least 5 training epochs. Herein an artificial olfactory chemical‐resistant synapse consisting of 3D hierarchical WO3@WO3 nanofibers are demonstrated. The nanofibers exhibit persistent resistance responses through chemical exposures due to the strong chemisorption of water molecules. Typical synaptic behaviors including paired‐pulse facilitation, long‐term −1 short‐term memory, and learning experience have been achieved. Next, a recurrent neural network that is committed to processing the time‐dependent data is used to identify gas‐phase chemicals of 3‐hydroxy‐2‐butanone, triethylamine, and trimethylamine. Training‐free gas recognition has been realized by a WO3@WO3 nanofiber synapse only, in which the accuracy is above 90% at the first epoch. The results have great potential to satisfy stringent performance requirements on artificial perception systems.

A Novel Underwater Acoustic Target Recognition Method Based on MFCC and RACNN.

In ocean remote sensing missions, recognizing an underwater acoustic target is a crucial technology for conducting marine biological surveys, ocean explorations, and other scientific activities that take place in water. The complex acoustic propagation characteristics present significant challenges for the recognition of underwater acoustic targets (UATR). Methods such as extracting the DEMON spectrum of a signal and inputting it into an artificial neural network for recognition, and fusing the multidimensional features of a signal for recognition, have been proposed. However, there is still room for improvement in terms of noise immunity, improved computational performance, and reduced reliance on specialized knowledge. In this article, we propose the Residual Attentional Convolutional Neural Network (RACNN), a convolutional neural network that quickly and accurately recognize the type of ship-radiated noise. This network is capable of extracting internal features of Mel Frequency Cepstral Coefficients (MFCC) of the underwater ship-radiated noise. Experimental results demonstrate that the proposed model achieves an overall accuracy of 99.34% on the ShipsEar dataset, surpassing conventional recognition methods and other deep learning models.

A Light-Weight Artificial Neural Network for Recognition of Activities of Daily Living

Human activity recognition (HAR) is essential for the development of robots to assist humans in daily activities. HAR is required to be accurate, fast and suitable for low-cost wearable devices to ensure portable and safe assistance. Current computational methods can achieve accurate recognition results but tend to be computationally expensive, making them unsuitable for the development of wearable robots in terms of speed and processing power. This paper proposes a light-weight architecture for recognition of activities using five inertial measurement units and four goniometers attached to the lower limb. First, a systematic extraction of time-domain features from wearable sensor data is performed. Second, a small high-speed artificial neural network and line search method for cost function optimization are used for activity recognition. The proposed method is systematically validated using a large dataset composed of wearable sensor data from seven activities (sitting, standing, walking, stair ascent/descent, ramp ascent/descent) associated with eight healthy subjects. The accuracy and speed results are compared against methods commonly used for activity recognition including deep neural networks, convolutional neural networks, long short-term memory and convolutional–long short-term memory hybrid networks. The experiments demonstrate that the light-weight architecture can achieve a high recognition accuracy of 98.60%, 93.10% and 84.77% for seen data from seen subjects, unseen data from seen subjects and unseen data from unseen subjects, respectively, and an inference time of 85 μs. The results show that the proposed approach can perform accurate and fast activity recognition with a reduced computational complexity suitable for the development of portable assistive devices.

Molecular mechanism of the role of Mare Nectaris in the Feng-Flavor Baijiu aging

A special storage container called Mare Nectaris bears a close relationship with Feng-flavor Baijiu quality during aging. The goal of this work is to unveil molecular mechanism of the role of Mare Nectaris in the Feng-flavor Baijiu aging as well as establishes a classification method of vintage Baijiu. Based on ultra high performance liquid chromatography quadrupole-orbitaltrap high resolution mass spectrometry (UHPLC-Q-Orbitrap HRMS), most micromolecules especially nonvolatile molecules in Baijiu can be accurately identified and in total of 45 characteristic molecules were discovered through application of the principal component analysis (PCA), partial least square-discriminant analysis (PLS-DA), variable importance in projection (VIP). With the increase of aging year, in addition to 5-fluoro-2,3,4-three chlorobenzoic acid decreased, the other 44 components showed remarkable up-regulation, concretely comprising 10 kinds of esters, 6 organic acids, 8 amino acids, 11 bioactive components and 9 other components. By utilizing artificial neural network recognition model which used 45 characteristic molecules as inputs and 8 groups various aging years as outputs, the prediction accuracy of Baijiu vintage was up to 100%.

A novel application of artificial neural network for recognition of target behind the wall

AbstractIn this article, a methodology for recognition of an object behind a wall has been proposed. So far research is mainly focused on target detection and localization issue. Very limited work has been reported so far for complete identification (ie, shape and size) of the target. Therefore in this article, a novel Artificial Neural Network based technique is proposed for identification of the targets behind the wall with the through the wall imaging system. For this purpose, an active microwave imaging radar system operating at frequency 3.5‐5.5 GHz has been ingeniously designed. An extensive number of experiments were performed while keeping targets made of metal and wood in different shapes and sizes. For identification of the target, an artificial neural network based image reconstruction model has been proposed. From experimental results, it was found that once training of proposed neural network with sample teaching images has been done, it could automatically rectify distortions and reconstruct nearly actual shape and size of the target for test targets. The obtained results are quite satisfactory.

OPTICAL FEEDBACK BASED ON THE PHOTOMETRYBY ELLIPSOIDAL REFLECTOR IN BIONIC FINGERS APPLICATION

Background. The investigation focus is the specificity of effective recognition of surface types during the interaction between finger prosthesis and manipulation objects, and organization of optical feedback system in the control module of bionic limb.Objective. The purpose of the paper is development and testing of method of optical feedback organization in the systems of bionic prostheses of human hand fingers.Methods. The developed system of optical feedback is based on the microcontroller measuring device with infrared (IR) optical emitter and sensor, ellipsoidal reflector, and artificial neural network.Results. During the research the comparison of application effectivity of optical feedback system with ellipsoidal reflector and without one was performed. The analysis was performed by classification of set with twelve, eleven, and ten types of surfaces of investigated specimens with the means of artificial neural network. The obtained accuracy of surface recognition for the system without an ellipsoidal reflector was 77%, 82% and 87%. In turn, the accuracy of surface recognition in the application of the ellipsoidal reflector was 94%, 98%, and 100%, correspondingly. Such results prove the possibility of further use of the photometry system by ellipsoidal reflectors for organization of constructive parts or complete feedback modules of the bionic fingers.Conclusions. The organization system of optical feedback based on optical emitter and sensor, ellipsoidal reflectors and artificial neural network for recognition of kinds of separate surfaces, with which can interact bionic prostheses fingers is proposed. The proposed system proved the certainty in recognition of limited set of investigated specimens. The efficiency of such system can be improved by using sensors array and wider data set for training.

Influence of denoising of input signal on classification of object location by artificial neural network in ultrawideband radiointroscopy

Background: A wide usage of impulse ultrawideband subsurface radars for a number of practical approaches in archeology, construction and humanitarian demining is holding back because of presence of noises and clutters of high level in the reflected field. It often makes the object classification practically unreal for at not big depths and distances from receiving and transmitting antennas. Besides of using special antenna system designs to improve recognition results, it is interesting to apply modern digital signal filtering techniques. Objectives: To investigate the influence of denoising on the quality of artificial neural network recognition of subsurface objects and their coordinates for a model of additive gaussian noise of a different noise level. Materials and methods: In this paper the idea of improving the stability of recognition of hidden objects in the presence of outside noise by previous processing of input signals with the latest popular noise reduction methods, such as the caterpillar method and wavelet transform method is verified. To eliminate the randomness of the result of the neural network response for each realization of the additive noise of a given level, a sufficient number of attempts are calculated for each of the methods, and statistics are provided to illustrate the effectiveness of each of the approaches. To check the hypothesis of the efficiency of input signal denoising the numerical simulation of the model of a real ground surface with subsurface object is carried out by means of Finite Difference Time Domain method (FDTD). The artificial neural network is trained on the obtained ideal time dependences of the amplitudes of the reflected field to correctly recognize the position of the object. The training is subsequently checked on the same input signals with additional noise of a certain level. Recognition errors in the last case are compared with similar errors when popular noise reduction procedures are applied to noisy input signals. Results: It is demonstrated that artificial neural networks have good approximating properties capable to effectively resist the noises in the input signals It is shown that for all noise levels, the caterpillar method statistically degrades the quality of an object recognition. The wavelet-transform method statistically improves slightly the classification of objects than for absence of denoising, but this result is not stable. Conclusion: For effective application of methods of noise filtration in received signals of impulse radar it is nessusary to have previous knowledge about noise character or peculiarities of useful signal. Implementation of denoising techniques without the use of this knowledge cannot improve the recognition quality of surface objects.

Development of robototechnological complex of intellectual management by bread manufacturing for technological loading territories

The object of research is a technological line for production of bakery products. One of the most problematic places in the technological process of industrial production of bakery products is the incoming quality control and preliminary preparation of ingredients, as well as the implementation of the processes of preparation of dough and kneading dough. To improve and intensify the relevant processes, it is proposed to use robotic technological complexes with an intelligent control system and built-in intensifiers. The principle of the intensifiers of the technological process is based on ultrasonic technologies of water treatment, mixing of components, intensification of physical and chemical processes of bread production, quality control of the process. For the production of bread with therapeutic and prophylactic characteristics, the controlled ultrasonic cavitation effects are studied. It has been proved that processes of ultrasonic cavitation, disintegration, coagulation of water, flour, oil, liquid yeast, hydrochloric and sugar solutions and other ingredients promote the production of medical type. Ultrasonic cavitation causes increased fluid mixing of the micro streams that form around the oscillate bubbles. Such mixing is especially useful in the technological process of producing bread for dispersion when enriched with vitamins and minerals. Systems for automated control of parameters of pre-dough-dough based on high-frequency ultrasonic vibrations are developed. It is proved that robotic technological complexes with developed sensor system allow to reduce the salt and sugar content in bread by 15–20 % without changing the taste of the product. A multi-level intelligent system of automated management of the technological process of bread production has been developed. The architecture of this system uses a robotic technology complex with: – intellectual decision support system and training blocks; – databases and knowledge; – block output information on the corporate performance monitor; – automated workplace with an interface system; – artificial neural network for recognition of emergency, abnormal and normal situations. Thus, with the help of the robotic technological complexes built into the technological process of bread production, it is possible to achieve a higher quality of bakery products.

Using Artificial Neural Networks for Recognition of Control Chart Pattern

Using Artificial Neural Networks for Recognition of Control Chart Pattern

Film Thickness Measurement Of Mechanical Seal Based On Cascaded Artificial Neural Network Recognition Model

Abstract Mechanical seal end faces are separated by a thin fluid film. The thickness of this film must be optimized so as to preventing the serious friction of two end faces and minimizing the fluid leakage. The micro scope condition monitoring of end face is of importance to ensure mechanical seals run normally. A method for measuring the film thickness of end faces and detecting the friction of end faces of mechanical seal has been presented in this paper. Eddy current sensors embedded in the stationary ring of mechanical seals are used to directly measure the thickness of the liquid-lubricated film. The Eddy current signal is decomposed by empirical mode decomposition into a series of intrinsic mode function. The information reflecting the film thickness is obtained by eliminating the false intrinsic mode function components. Acoustic emission sensor placed on the lateral of stationary ring is used to detect the friction of end faces. In order to decrease the acoustic emission signal’s noise, wavelet packet and kernel principal component analysis are used to extract the data features. Then cascaded decision is presented to improve the recognition rate of artificial neural network, by which the film thickness can be estimated accurately. With a set of tests, the results demonstrate that the method is effective. It can be widely used to take measurement of the film thickness in industrial field.

An Improved Artificial Neural Network for Recognition of Beam Pump Dynamometer Card

Beam pump dynamometer card plays an important role in identifying the production state of oil wells. With an ability to reflect any non-linear mapping relationship, the artificial neural network (ANN) can be used in pattern recognition. This paper illuminates ANN realization in identifying fault kinds of dynamometer cards, including a back-propagation algorithm, characteristics of the Dynamometer card and some examples. It is concluded that the buildup of a neural network and the abstract of dynamometer cards are important to successful application.

Discrete Wavelet Transforms and Artificial Neural Networks for Recognition of Isolated Spoken Words

recognition is a fascinating application of Digital Signal Processing and has many real-world applications. In this paper, a speech recognition system is developed for isolated spoken words using Discrete Wavelet Transforms (DWT) and Artificial Neural Networks (ANN). Speech signals are one-dimensional and are random in nature. Isolated words from Malayalam, one of the four major Dravidian languages of southern India are chosen for recognition. Daubechies wavelets are employed here. A multi- layer neural network trained with back propagation training algorithm is used for classification purpose. The proposed method is implemented for 50 speakers uttering 20 isolated words each. The experimental results show good recognition accuracy and the efficiency of combining these two techniques.

Spatially resolved mapping of disorder type and distribution in random systems using artificial neural network recognition

The spatial variability of the polarization dynamics in thin film ferroelectric capacitors was probed by recognition analysis of spatially resolved spectroscopic data. Switching spectroscopy piezoresponse force microscopy (SSPFM) was used to measure local hysteresis loops and map them on a two dimensional (2D) random-bond, random-field Ising model. A neural-network based recognition approach was utilized to analyze the hysteresis loops and their spatial variability. Strong variability is observed in the polarization dynamics around macroscopic cracks because of the modified local-elastic and electric-boundary conditions, with the most pronounced effect on the length scale of \ensuremath{\sim}100 nm away from the crack. The recognition approach developed here is universal and can potentially be applied for arbitrary macroscopic and spatially resolved data, including temperature- and field-dependent hysteresis, I-V curve mapping, electron microscopy electron energy loss spectroscopy (EELS) imaging, and many others.

Industrial Sewage Evaluation Based on Fuzzy Neural Network

Abstract The purpose of this paper is to propose a new method of industrial sewage evaluation. A comprehensive evaluation is presented to industrial sewage on the base of fuzzy recognition firstly. Then a model of fuzzy artificial neural network recognition is set up to evaluate industrial sewage. Finally, the feasibility and the effectiveness of approaches of industrial sewage evaluation based on fuzzy neural network are validated by some examples of industrial sewage evaluation of Zigong City here.

Recognition of Individual Holstein Cattle by Imaging Body Patterns

A computer vision system was designed and validated to recognize an individual Holstein cattle by processing images of their body patterns. This system involves image capture, image pre-processing, algorithm processing, and an artificial neura ln etwork recognition algorithm. Optimum management of individuals is one of the most important factors in keeping cattle healthy and productive. In this study, an image-processing system was used to recognize individual Holstein cattle by identifying the body-pattern images captured by a charge-coupled device (CCD). A recognition system was developed and applied to acquire images of 49 cattles. The pixel values of the body images were transformed into input data comprising binary signals for the neural network. Images of the 49 cattle were analyzed to learn input layer elements, and ten cattles were used to verify the output layer elements in the neural network by using an individual recognition program. The system proved to be reliable for the individual recognition of cattles in natural light. (Asian-Aust. J. Anim. Sci. 2005. Vol 18, No. 8: 1194-1198)

Efficient control chart pattern recognition through synergistic and distributed artificial neural networks

Accurate and fast control chart pattern recognition is essential for efficient system monitoring to maintain the production of high-quality goods. This paper addresses three major issues of control chart pattern recognition: (a) transparency, (b) accuracy and (c) fast detection of abnormal patterns. A new approach is described which uses novel shape features extracted from a control chart pattern (CCP) instead of the unprocessed CCP data or its statistical properties. These features represent the shape of the CCP explicitly. A set of algorithms is described for extraction of the shape features from a CCP. The paper discusses the use of artificial neural networks for recognition of the shape features. Synergistic, distributed and distributed synergistic neural networks are proposed for learning efficiently non-linear characteristics and overlapping ranges of values of the data set describing CCPs. The paper presents the results of analysing several hundred control chart patterns and gives a comparison with those reported in previous work.

Artificial neural networks for recognition of electrocardiographic lead reversal

Misplacement of electrodes during the recording of an electrocardiogram (ECG) can cause an incorrect interpretation, misdiagnosis, and subsequent lack of proper treatment. The purpose of this study was twofold: (1) to develop artificial neural networks that yield peak sensitivity for the recognition of right/left arm lead reversal at a very high specificity; and (2) to compare the performances of the networks with those of 2 widely used rule-based interpretation programs. The study was based on 11,009 ECGs recorded in patients at an emergency department using computerized electrocardiographs. Each of the ECGs was used to computationally generate an ECG with right/left arm lead reversal. Neural networks were trained to detect ECGs with right/left arm lead reversal. Different networks and rule-based criteria were used depending on the presence or absence of P waves. The networks and the criteria all showed a very high specificity (99.87% to 100%). The neural networks performed better than the rule-based criteria, both when P waves were present (sensitivity 99.1%) or absent (sensitivity 94.5%). The corresponding sensitivities for the best criteria were 93.9% and 39.3%, respectively. An estimated 300 million ECGs are recorded annually in the world. The majority of these recordings are performed using computerized electrocardiographs, which include algorithms for detection of right/left arm lead reversals. In this study, neural networks performed better than conventional algorithms and the differences in sensitivity could result in 100,000 to 400,000 right/left arm lead reversals being detected by networks but not by conventional interpretation programs.

Artificial neural networks for recognition of 3-d partial discharge patterns

Partial discharge (PD) measurements have been carried out over the years to assess insulation systems in power apparatus for their integrity and design deficiencies. Digital PD recording, processing and its presentation as 3-d patterns are recent trends in both industry and testing laboratories. Interpretation of these patterns can lead to evaluation of the cause of PD. A need arises to look for methods in the domain of pattern recognition for automating this process. In this context, this paper presents results to demonstrate the possibility of using pattern recognition capabilities offered by a multilayer neural network to recognize 3-d PD patterns. >