Recognition Subsystem Research Articles

Procedure for describing traffic situation scene development

PurposeThis study aims to comprehensively address the challenge of delineating traffic scenarios in video footage captured by an embedded camera within an autonomous vehicle.Design/methodology/approachThis methodology involves systematically elucidating the traffic context by leveraging data from the object recognition subsystem embedded in vehicular road infrastructure. A knowledge base containing production rules and logical inference mechanism was developed. These components enable real-time procedures for describing traffic situations.FindingsThe production rule system focuses on semantically modeling entities that are categorized as traffic lights and road signs. The effectiveness of the methodology was tested experimentally using diverse image datasets representing various meteorological conditions. A thorough analysis of the results was conducted, which opens avenues for future research.Originality/valueOriginality lies in the potential integration of the developed methodology into an autonomous vehicle’s control system, working alongside other procedures that analyze the current situation. These applications extend to driver assistance systems, harmonized with augmented reality technology, and enhance human decision-making processes.

Required Operational Capabilities of the Recognition Subsystem in the Anti-Access/Area Denial System

The subject of this article is the issue of "the required operational capabilities of the reconnaissance subsystem in the anti-access/area denial system".
 When defining the assumptions for the reconnaissance subsystem in the A2/AD system, several assumptions were adopted for the analyses resulting from the concept of military operations from the beginning of the 21st century. These assumptions were derived from analyses carried out as part of conferences, seminars, research work, and army command and staff exercises (see: Wrzosek, 2020). Moreover, epistemological achievements in the field of military thought were used to generate solutions.
 The accumulated knowledge is the basis for concluding that the current reconnaissance subsystem does not meet the requirements, full perception of various reconnaissance data obtained from the area of operations and the translation of aggregate knowledge into the cognitive sphere of commanders and staff. Additionally, it was established that in the current organizational and functional solutions, the commander's knowledge is based on an understanding of the area of operation and the enemy troops' situation. Meanwhile, the new scope of reconnaissance tasks under the A2/AD system requires collecting knowledge from all domains where the fate of a military conflict is decided.

Design and Implementation of an Artificial Intelligence of Things-Based Autonomous Mobile Robot System for Cleaning Garbage

This article proposes an autonomous mobile robot (AMR) system based on the artificial intelligence of things (AIoT) for collecting garbage. The proposed system consists of an AMR subsystem, a robot operating system (ROS) robot subsystem, an AI garbage recognition subsystem, solar power supply subsystem, and a smart trash can subsystem. The AMR subsystem uses a motor drive, which is built on an aluminum alloy metal base and is connected to an AI edge computing module via an app, to control the robot. The ROS subsystem is built on an edge computing module, which is coupled with the motor drive module to control the robot arm and the robot gripper, and is connected to the AI edge computing module to control the robot arm through the app. The AI garbage recognition subsystem recognizes the types of waste in real time through an image sensor (camera). This subsystem adopts multiple sensors that can monitor the level of waste in the trash can. Moreover, a map on a web platform locates all trash cans. Servomotors enable automatic opening and closing of the lid to prevent the unpleasant smell of garbage from wafting. A user connects to the AI edge computing module through an app to monitor the robot and control the movement of the robot to any location on the map.

Helix Project: Exploring the Social Internet of Things (SIoT) in Care of Blind People

The wide variety of available mobile devices, associated with the increase in wireless communication network availability, enables an increasing offer of mobile and ubiquitous technologies. In this sense, this article proposes an approach called Helix-SIoT, which explores the Social Internet of Things concepts to support Blind People (BP). The objective is to enhance the treatment of the different sensors that integrate the assistance of these people with SIoT use. For this, the proposal employs Context Awareness and Multiple Criteria Decision Analysis, helping people's autonomy in terms of mobility, response times, and the choice of the Caregiver in the most transparent way possible. The functional architecture of the Helix approach operates integrated with the Adaptation and Context Recognition Subsystem of the EXEHDA middleware, which provides support for the acquisition, storage, and processing of context information used by the mobile Helix-SIoT applications.

Design of an intelligent vibration screening system for armyworm pupae based on image recognition

Biological control is a green and effective pest management method to mitigate or inhibit pests and pest effects by introducing natural enemies. The stinkbug is one of the insects used for biological control and fed on armyworm. It is necessary to separate the farm-raised armyworm from the sticky substrate after pupation. At present, the work is done manually, with high labor intensity, low efficiency and high cost. In this paper, an intelligent vibration screening system of armyworm pupae based on image recognition is developed. The system can effectively separate the pupae from the sticky substrate without damaging the pupae. The system is comprised of vibration subsystem, separation subsystem, image recognition subsystem and control subsystem. The pupa material is sticky and cannot be shaken off by vibration alone. Therefore, the separation subsystem uses the reciprocating motion of the brush to separate the armyworm from the substrate. Then, the image recognition technology is used to adjust the feed quantity in real time to ensure the optimal screening effect. Finally, the optimal working parameters, including separation speed, vibration frequency and feed quantity, are determined by using the design method of linear regression and linear programming. The experimental results show that the screening rate of the screening system can reach 90%, and the damage rate can be controlled within 5%.

HMM speech recognition study of an Improved Particle Swarm Optimization Based on Self-Adaptive Escape (AEPSO)

Aiming at the single evaluation content of the current train simulation training system, the process of voice interaction in the daily work of train conductors cannot be effectively and comprehensively reflected, and the coordination between dispatching terminology and train running equipment are not considered. In order to realize the objective evaluation of whether the terms are clear and the semantics accurate and the consistency check of “schedule terms-equipment operation”, the HMM speech recognition subsystem of an Improved Particle Swarm Optimization Based on Self-Adaptive Escape (AEPSO) for railway dispatching was constructed. Compared with the traditional Baum-Welch algorithm, the experiment result shows that AEPSO-BW algorithm has certain advantages in speech recognition compared with traditional Baum-Welch algorithm. It provides a more comprehensive evaluation index for the training and assessment of vehicle service personnel.

Room-localized speech activity detection in multi-microphone smart homes

Voice-enabled interaction systems in domestic environments have attracted significant interest recently, being the focus of smart home research projects and commercial voice assistant home devices. Within the multi-module pipelines of such systems, speech activity detection (SAD) constitutes a crucial component, providing input to their activation and speech recognition subsystems. In typical multi-room domestic environments, SAD may also convey spatial intelligence to the interaction, in addition to its traditional temporal segmentation output, by assigning speech activity at the room level. Such room-localized SAD can, for example, disambiguate user command referents, allow localized system feedback, and enable parallel voice interaction sessions by multiple subjects in different rooms. In this paper, we investigate a room-localized SAD system for smart homes equipped with multiple microphones distributed in multiple rooms, significantly extending our earlier work. The system employs a two-stage algorithm, incorporating a set of hand-crafted features specially designed to discriminate room-inside vs. room-outside speech at its second stage, refining SAD hypotheses obtained at its first stage by traditional statistical modeling and acoustic front-end processing. Both algorithmic stages exploit multi-microphone information, combining it at the signal, feature, or decision level. The proposed approach is extensively evaluated on both simulated and real data recorded in a multi-room, multi-microphone smart home, significantly outperforming alternative baselines. Further, it remains robust to reduced microphone setups, while also comparing favorably to deep learning-based alternatives.

A Multimodal Approach to Enhancing Automobile Security

The continuous increase in use of the automobiles has resulted in increase in vehicular traffic and social vices such as theft, thus necessitating efforts for improving automobile security. As a result of this, various solutions such as satellite monitoring and General Packet Radio Services have been adopted to track vehicle locations over time. This article combines the uniqueness of an automobile plate number and the face(s) of an authorized vehicle driver to enhance the automobile security system, hence eliminating the inefficiency of the existing tag system while reducing the cases of unauthorized automobile use and outright theft. A weighted score approach was used in evaluating the efficiency of the character recognition subsystem with an accuracy of 89.3% recorded. The face recognition subsystem also achieved reliable genuine and an imposter attempt at 0.0 confidence level. The designed system ensured security of an automobile through the combination of a plate number and image recognition, thereby preventing an unauthorized driver from taking a vehicle out of a controlled area.

Efficiency analysis of dynamic object detection in computer vision system

The concept of computer vision and its key aspects are given. Problems of dynamic objects detection are considered. The latest approaches are described to improve the accuracy and speed of recognition. Their main advantages and disadvantages are considered. The structure of the computer vision system and the interaction of its components are shown. A method for embedding a motion detection module to increase the efficiency of the recognition subsystem is proposed. A motion vector search algorithm is described, the main idea of which is to divide a frame into blocks and search for similar parts in adjacent frames. Experimental results showing the effectiveness of the approach are demonstrated.

Gesture recognition system based on Convolutional neural networks

This article tells about gesture recognition system based on Convolutional neural nets. System consists of two parts: the tracking and detection subsystem and gesture recognition subsystem.

The role of speech technology in biometrics, forensics and man-machine interface

<span lang="EN-US">Day by day Optimism is growing that in the near future our society will witness the Man-Machine Interface (MMI) using voice technology. Computer manufacturers are building voice recognition sub-systems in their new product lines. Although, speech technology based MMI technique is widely used before, needs to gather and apply the deep knowledge of spoken language and performance during the electronic machine-based interaction. Biometric recognition refers to a system that is able to identify individuals based on their own behavior and biological characteristics. Fingerprint success in forensic science and law enforcement applications with growing concerns relating to border control, banking access fraud, machine access control and IT security, there has been great interest in the use of fingerprints and other biological symptoms for the automatic recognition. It is not surprising to see that the application of biometric systems is playing an important role in all areas of our society. Biometric applications include access to smartphone security, mobile payment, the international border, national citizen register and reserve facilities. The use of MMI by speech technology, which includes automated speech/speaker recognition and natural language processing, has the significant impact on all existing businesses based on personal computer applications. With the help of powerful and affordable microprocessors and artificial intelligence algorithms, the human being can talk to the machine to drive and control all computer-based applications. Today's applications show a small preview of a rich future for MMI based on voice technology, which will ultimately replace the keyboard and mouse with the microphone for easy access and make the machine more intelligent.</span>

Traffic Sign Detection and Recognition for Driving Assistance System

In this paper, we present a traffic sign detection and recognition for a driving assistance system. The proposed approach consists of two subsystems for detection and recognition. First, the road sign detection subsystem adopts the color information to filter out most of irrelevant image regions. Image segmentation and hierarchical grouping are then used to select candidate regions of road signs. For the road sign recognition subsystem, a convolutional neural network (CNN) is adopted to classify traffic signs for candidate regions. Experimental results show that our approach can obtain the desired results effectively.

Investigation of Combined use of MFCC and LPC Features in Speech Recognition Systems

Statement of the automatic speech recognition problem, the assignment of speech recognition and the application fields are shown in the paper. At the same time as Azerbaijan speech, the establishment principles of speech recognition system and the problems arising in the system are investigated. The computing algorithms of speech features, being the main part of speech recognition system, are analyzed. From this point of view, the determination algorithms of Mel Frequency Cepstral Coefficients (MFCC) and Linear Predictive Coding (LPC) coefficients expressing the basic speech features are developed. Combined use of cepstrals of MFCC and LPC in speech recognition system is suggested to improve the reliability of speech recognition system. To this end, the recognition system is divided into MFCC and LPC-based recognition subsystems. The training and recognition processes are realized in both subsystems separately, and recognition system gets the decision being the same results of each subsystems. This results in decrease of error rate during recognition. The training and recognition processes are realized by artificial neural networks in the automatic speech recognition system. The neural networks are trained by the conjugate gradient method. In the paper the problems observed by the number of speech features at training the neural networks of MFCC and LPC-based speech recognition subsystems are investigated. The variety of results of neural networks trained from different initial points in training process is analyzed. Methodology of combined use of neural networks trained from different initial points in speech recognition system is suggested to improve the reliability of recognition system and increase the recognition quality, and obtained practical results are shown.

Human robot interaction for manipulation tasks based on stroke gesture recognition

PurposeThe purpose of this paper is to extend the usage of stroke gestures in manipulation tasks to make the interaction between human and robot more efficient.Design/methodology/approachIn this paper, a set of stroke gestures is designed for typical manipulation tasks. A gesture recognition and parameter extraction system is proposed to exploit the information in stroke gestures drawn by the users.FindingsThe results show that the designed gesture recognition subsystem can reach a recognition accuracy of 99.00 per cent. The parameter extraction subsystem can successfully extract parameters needed for typical manipulation tasks with a success rate about 86.30 per cent. The system shows an acceptable performance in the experiments.Practical implicationsUsing stroke gesture in manipulation tasks can make the transmission of human intentions to the robots more efficient. The proposed gesture recognition subsystem is based on convolutional neural network which is robust to different input. The parameter extraction subsystem can extract the spatial information encoded in stroke gestures.Originality/valueThe author designs stroke gestures for manipulation tasks which is an extension of the usage of stroke gestures. The proposed gesture recognition and parameter extraction system can make use of stroke gestures to get the type of the task and important parameters for the task simultaneously.

Decision-level fusion for single-view gait recognition with various carrying and clothing conditions

Gait recognition is one of the latest and attractive biometric techniques, due to its potential in identification of individuals at a distance, unobtrusively and even using low resolution images. In this paper we focus on single lateral view gait recognition with various carrying and clothing conditions. Such a system is needed in access control applications whereby a single view is imposed by the system setup. The gait data is firstly processed using three gait representation methods as the features sources; Accumulated Prediction Image (API) and two new gait representations namely; Accumulated Flow Image (AFI) and Edge-Masked Active Energy Image (EMAEI). Secondly, each of these methods is tested using three matching classification schemes; image projection with Linear Discriminant Functions (LDF), Multilinear Principal Component Analysis (MPCA) with K-Nearest Neighbor (KNN) classifier and the third method: MPCA plus Linear Discriminant Analysis (MPCA+LDA) with KNN classifier. Gait samples are fed into the MPCA and MPCALDA algorithms using a novel tensor-based form of the gait images. This arrangement results into nine recognition sub-systems. Decisions from the nine classifiers are fused using decision-level (majority voting) scheme. A comparison between unweighted and weighted voting schemes is also presented. The methods are evaluated on CASIA B Dataset using four different experimental setups, and on OU-ISIR Dataset B using two different setups. The experimental results show that the classification accuracy of the proposed methods is encouraging and outperforms several state-of-the-art gait recognition approaches reported in the literature.

Enhancing the Quality of Rice Milling by Systematic Innovation Techniques

Rice is a key food of humankind, the third largest food crop in the world, behind only corn and wheat. The process of removing the rice husk, the rice skin portion, and the rice germ to make white rice is called rice milling. This empirical study applied some tools (Cause-and-Effect Diagram, QFD, FAA and TRIZ) to improve the quality of rice milling. At first, this study used Yixing Rice Milling Company in Taiwan to illustrate the process of fully automated rice milling. There are eight steps in the rice milling process system: testing, drying, refrigeration, hulling, milled rice, rice washing, color sorting, and packaging. Second, using risk assessment theory, the eight steps of the process were evaluated to determine which steps were the most vulnerable to risks. Four key factors in poor milling quality were identified: the hue recognition subsystem, the cold storage structure, ventilation, and the air pressure nozzle subsystem. Finally, the four key factors were improved with systematic innovation methods (TRIZ). We then made recommendations for improving the quality of rice production operations to Yixing Rice Milling Company.

An Approach to Ballet Dance Training through MS Kinect and Visualization in a CAVE Virtual Reality Environment

This article proposes a novel framework for the real-time capture, assessment, and visualization of ballet dance movements as performed by a student in an instructional, virtual reality (VR) setting. The acquisition of human movement data is facilitated by skeletal joint tracking captured using the popular Microsoft (MS) Kinect camera system, while instruction and performance evaluation are provided in the form of 3D visualizations and feedback through a CAVE virtual environment, in which the student is fully immersed. The proposed framework is based on the unsupervised parsing of ballet dance movement into a structuredposture spaceusing the spherical self-organizing map (SSOM). A unique feature descriptor is proposed to more appropriately reflect the subtleties of ballet dance movements, which are represented asgesture trajectoriesthrough posture space on the SSOM. This recognition subsystem is used to identify the category of movement the student is attempting when prompted (by a virtual instructor) to perform a particular dance sequence. The dance sequence is then segmented and cross-referenced against a library of gestural components performed by the teacher. This facilitates alignment and score-based assessment of individual movements within the context of the dance sequence. An immersive interface enables the student to review his or her performance from a number of vantage points, each providing a unique perspective and spatial context suggestive of how the student might make improvements in training. An evaluation of the recognition and virtual feedback systems is presented.

Subsystem of mass image recognition of archival documents

The description of main features and problems of mass recognition of archival documents is presented in the beginning of the paper. Key design problems are considered, the examples are shown and different kinds of solutions are advised. After that, the scheme of electronic archive and the scheme of recognition subsystem are shown with the description of their components, functions and business processes.

Remote Fault Diagnosis System Based on EMD and SVM for Heavy Rolling-Mills

Aiming at the requirements of fault diagnosis for heavy mills, the architecture of remote fault diagnosis system based on EMD and SVM for heavy mills is built. Then the function of 3 main subsystems in the prototype system is introduced: Pattern recognition subsystem is used to evaluate healthy state of equipment with SVM classification algorithm; Fault location subsystem is used to fix fault position in the equipment with the method of Empirical Mode Decomposition and Hilbert-Huang Transform; Remaining life prediction subsystem is used to make a prediction of equipments health trends with SVM regression algorithm. At last, a remote fault diagnosis system based on website is established.

Sensor-based surgical activity recognition in unconstrained environments

Introduction: Automatic surgical activity recognition in the operating room (OR) is mandatory to enable assistive surgical systems to manage the information presented to the surgical team. Therefore the purpose of our study was to develop and evaluate an activity recognition model. Material and methods: The system was conceived as a hierarchical recognition model which separated the recognition task into activity aspects. The concept used radio frequency identification (RFID) for instrument recognition and accelerometers to infer the performed surgical action. Activity recognition was done by combining intermediate results of the aspect recognition. A basic scheme of signal feature generation, clustering and sequence learning was replicated in all recognition subsystems. Hidden Markov models (HMM) were used to generate probability distributions over aspects and activities. Simulated functional endoscopic sinus surgeries (FESS) were used to evaluate the system. Results and discussion: The system was able to detect surgical activities with an accuracy of 95%. Instrument recognition performed best with 99% accuracy. Action recognition showed lower accuracies with 81% due to the high variability of surgical motions. All stages of the recognition scheme were evaluated. The model allows distinguishing several surgical activities in an unconstrained surgical environment. Future improvements could push activity recognition even further.