Obtained Recognition Rate Research Articles

Enhancing person re-identification on RGB-D data with noise free pose-regularized color and skeleton distance features

Noisy features may introduce irrelevant or incorrect features that can lead to incorrect classifications and lower accuracy. This can be especially problematic in tasks such as person re-identification (ReID), where subtle differences between individuals need to be accurately captured and distinguished. However, the existing ReID methods directly use noisy and limited multimodality features for similarity measures. It is crucial to use robust features and pre-processing techniques to reduce the effects of noise and ensure accurate classification. As a solution, we employ a Gaussian filter to eliminate the Gaussian noise from RGB-D data in the pre-processing stage. For similarity measure, the color descriptors are computed using the top eight peaks of the 2D histogram constructed from pose regularized partition grid cells, and eleven different skeleton distances are considered. The proposed method is evaluated on the BIWI RGBD-ID dataset, which comprises still (front view images) and walking set (images with varied pose and viewpoint) images. The obtained recognition rates of 99.15% and 94% on still and walking set images demonstrate the effectiveness of the proposed approach for the ReID task in the presence of pose and viewpoint variations. Further, the method is evaluated on and RGBD-ID and achieved improved performance over the existing techniques.

Comparing Speaker Adaptation Methods for Visual Speech Recognition for Continuous Spanish

Visual speech recognition (VSR) is a challenging task that aims to interpret speech based solely on lip movements. However, although remarkable results have recently been reached in the field, this task remains an open research problem due to different challenges, such as visual ambiguities, the intra-personal variability among speakers, and the complex modeling of silence. Nonetheless, these challenges can be alleviated when the task is approached from a speaker-dependent perspective. Our work focuses on the adaptation of end-to-end VSR systems to a specific speaker. Hence, we propose two different adaptation methods based on the conventional fine-tuning technique, the so-called Adapters. We conduct a comparative study in terms of performance while considering different deployment aspects such as training time and storage cost. Results on the Spanish LIP-RTVE database show that both methods are able to obtain recognition rates comparable to the state of the art, even when only a limited amount of training data is available. Although it incurs a deterioration in performance, the Adapters-based method presents a more scalable and efficient solution, significantly reducing the training time and storage cost by up to 80%.

A Ternary Neural Network with Compressed Quantized Weight Matrix for Low Power Embedded Systems

In this paper, we propose a method of transforming a real-valued matrix to a ternary matrix with controllable sparsity. The sparsity of quantized weight matrices can be controlled by adjusting the threshold during the training and quantizing process. A 3-layer ternary neural network was trained with the MNIST dataset using the proposed adjustable dynamic threshold. The sparsity of the quantized weight matrices varied from 0.1 to 0.6 and the obtained recognition rate reduced from 91% to 88%. The sparse weight matrices were compressed by the compressed sparse row format to speed up the ternary neural network, which can be deployed on low-power embedded systems, such as the Raspberry Pi 3 board. The ternary neural network with the sparsity of quantized weight matrices of 0.1 is 4.24 times faster than the ternary neural network without compressing weight matrices. The ternary neural network is faster as the sparsity of quantized weight matrices increases. When the sparsity of the quantized weight matrices is as high as 0.6, the recognition rate degrades by 3%, however, the speed is 9.35 times the ternary neural network's without compressing quantized weight matrices. Ternary neural network work with compressed sparse matrices is feasible for low-cost, low-power embedded systems.

A Hybrid Approach for Automatic Licence Plate Recognition System

Background: Automatic licence plate recognition system is used for various applications such as traffic monitoring, toll collection, car parking, law enforcement. Objective: In this paper, a convolutional neural network and support vector machine-based licence plate recognition system is proposed. Method: Firstly, from the input image of the vehicle, the characters are extracted and segmented. Then features of the segmented characters are extracted. The extracted features are classified using convolutional neural networks and support vector machine for the final recognition of the licence plate. Results: The obtained recognition rate by the hybridization of the convolutional neural network and the support vector machine is 96.5%. The obtained results for the proposed hybrid automatic licence plate system is compared with three other automatic licence plate systems based on neural network, support vector machine, and convolutional neural network. Conclusion: The proposed hybrid ALPR system performs better than NN, SVM, and CNN based ALPR systems.

Hand Gesture Recognition for Sign Language Using 3DCNN

Recently, automatic hand gesture recognition has gained increasing importance for two principal reasons: the growth of the deaf and hearing-impaired population, and the development of vision-based applications and touchless control on ubiquitous devices. As hand gesture recognition is at the core of sign language analysis a robust hand gesture recognition system should consider both spatial and temporal features. Unfortunately, finding discriminative spatiotemporal descriptors for a hand gesture sequence is not a trivial task. In this study, we proposed an efficient deep convolutional neural networks approach for hand gesture recognition. The proposed approach employed transfer learning to beat the scarcity of a large labeled hand gesture dataset. We evaluated it using three gesture datasets from color videos: 40, 23, and 10 classes were used from these datasets. The approach obtained recognition rates of 98.12%, 100%, and 76.67% on the three datasets, respectively for the signer-dependent mode. For the signer-independent mode, it obtained recognition rates of 84.38%, 34.9%, and 70% on the three datasets, respectively.

Automated System for Grading Apples using Convolutional Neural Network

Fruit grading is a process that affect quality control and fruit-processing industries to meet the efficiency of its production and society. However, these industries have suffered from lack of standards in quality control, higher time of grading and low product output because of the use of manual methods. To meet the increasing demand of quality fruit products, fruit-processing industries must consider automating their fruit grading process. Several algorithms have been proposed over the years to achieve this purpose and their works were based on color, shape and inability to handle large dataset which resulted in slow recognition accuracy. To mitigate these flaws, we develop an automated system for grading and classification of apple using Convolutional Neural Network (CNN) used in image recognition and classification. Two models were developed from CNN using ResNet50 as its convolutional base, a process called transfer learning. The first model, the apple checker model (ACM) performs the recognition of the image with two output connections (apple and non-apple) while the apple grader model (AGM) does the classification of the image that has four output classes (spoiled, grade A, grade B & grade C) if the image is an apple. A comparison evaluation of both models were conducted and experimental results show that the ACM achieved a test accuracy of 100% while the AGM obtained recognition rate of 99.89%.The developed system may be employed in food processing industries and related life applications.

3D geometry-based face recognition in presence of eye and mouth occlusions

This study proposes a novel occlusions detection and restoration strategy. The aim is to success with 3D face recognition even when faces are partially occluded by external objects. The method, which relies on geometrical facial properties, is designed for managing two types of facial occlusions (eye and mouth occlusions due to hands). First occlusions are detected and (if present) classified, by considering their effects on the 3D points cloud. Then, the occluded regions are progressively removed, and finally, the non-occluded symmetrical regions are used to restore the missing information. After the restoration process, face recognition is performed relying on the restored facial information and on the localized landmarks. The landmarking methodology relies on derivatives and on 12 differential geometry descriptors. The discriminating features adopted for facial comparison include shape index histograms, Euclidean and geodetical distances between landmarks, facial curves, and nose volume. Obtained recognition rates, evaluated on the whole Bosphorus database and on our private dataset, ranging from 92.55 to 97.20% depending on the completeness of data.

Classification of Three Types of Walking Activities Regarding Stairs Using Plantar Pressure Sensors

In this letter, we propose a novel method for classifying ambulatory activities using eight plantar pressure sensors within smart shoes. Using these sensors, pressure data of participants can be collected regarding level walking, stair descent, and stair ascent. Analyzing patterns of the ambulatory activities, we present new features with which to describe the ambulatory activities. After selecting critical features, a multi-class support vector machine algorithm is applied to classify these activities. Applying the proposed method to the experimental database, we obtain recognition rates up to 95.2% after six steps.

Reconstruction-Free Action Inference from Compressive Imagers.

Persistent surveillance from camera networks, such as at parking lots, UAVs, etc., often results in large amounts of video data, resulting in significant challenges for inference in terms of storage, communication and computation. Compressive cameras have emerged as a potential solution to deal with the data deluge issues in such applications. However, inference tasks such as action recognition require high quality features which implies reconstructing the original video data. Much work in compressive sensing (CS) theory is geared towards solving the reconstruction problem, where state-of-the-art methods are computationally intensive and provide low-quality results at high compression rates. Thus, reconstruction-free methods for inference are much desired. In this paper, we propose reconstruction-free methods for action recognition from compressive cameras at high compression ratios of 100 and above. Recognizing actions directly from CS measurements requires features which are mostly nonlinear and thus not easily applicable. This leads us to search for such properties that are preserved in compressive measurements. To this end, we propose the use of spatio-temporal smashed filters, which are compressive domain versions of pixel-domain matched filters. We conduct experiments on publicly available databases and show that one can obtain recognition rates that are comparable to the oracle method in uncompressed setup, even for high compression ratios.

Full body movements recognition – unsupervised learning approach with heuristic R-GDL method

In computer systems that are used for actions recognition the human movements are often represented by three-dimensional coordinates of body joints that are tracked by motion capture hardware. The motivation of our research was to propose a novel method for automatic generation of knowledge base for syntactic Gesture Description Language (GDL) classifier by analyzing unsegmented data recordings of gestures. We have proposed novel unsupervised learning approach to deal with this task. Because this process seems to be reverse engineering to GDL approach, the learning algorithm we introduce in this paper, is called Revers-GDL (R-GDL). The R-GDL machine-learning approach for full-body movements recognition is a novel method of time-varying multidimensional signals classification. The description of R-GDL and its validation is our original and never before published achievement. The evaluation of R-GDL was performed with k-fold cross validation on large dataset that contains 770 complete movements samples of 9 gym exercises performed by 14 persons and compared with results from multivariate normally continuous density hidden Markov model classifier. Depending on exercise type GDL obtained recognition rate at the level of 100% to 91%.

Improving the Recognition of Faces using PCA and SVM Optimized by DWT

Automatic recognition of people has received much attention during the recent years due to its many applications in different fields such as law enforcement, security applications or video indexing.In this paper,Wavelet transform is used for preprocessing the images in order to handle bad illumination, Principle Component Analysis (PCA) is used to play a key role in feature extractor and the SVM were used for classification.Support Vector Machines (SVMs) have been recently proposed as a new classifier for pattern recognition. We illustrate the potential of SVMs on the Cambridge ORL Face database, which consists of 400 images of 40 individuals, containing quite a high degree of variability in expression, pose, and facial details. The SVMs that have been used included the Linear (LSVM), Polynomial (PSVM), and Radial Basis Function (RBFSVM) SVMs, we obtain recognition rates as high as 97,9 in ORL face database with polynomial kernel (PSVM).

Fusion of Inertial and Depth Sensor Data for Robust Hand Gesture Recognition

This paper presents the first attempt at fusing data from inertial and vision depth sensors within the framework of a hidden Markov model for the application of hand gesture recognition. The data fusion approach introduced in this paper is general purpose in the sense that it can be used for recognition of various body movements. It is shown that the fusion of data from the vision depth and inertial sensors act in a complementary manner leading to a more robust recognition outcome compared with the situations when each sensor is used individually on its own. The obtained recognition rates for the single hand gestures in the Microsoft MSR data set indicate that our fusion approach provides improved recognition in real-time and under realistic conditions.

Voice Pathology Assessment Systems for Dysphonic Patients: Detection, Classification, and Speech Recognition

ABSTRACTIn the past decade, much research has been done on automatic detection and classification of vocal fold disorders, and these tasks continue to require further investigation. The aim of this study is to develop systems that may help in diagnosing patients from their speech. The systems will perform voice disorder detection, classification of voice disorders, and digit recognition. To find the best system, we will compare the system performance when using different voice features. We are the first to explore the use relative spectral transform perceptual linear predictive (RASTA-PLP) feature for speech pathology. The speech samples used in most of the literature are sustained vowels, while the speech samples we worked on are words, which are more natural. To evaluate the performance of the developed system, we used a database containing five types of vocal fold disorders. The database includes a total of 142 speakers half of them were normal speakers. The best accuracy achieved for the voice disorder detection system was 92.40%. In the voice disorder classification system, the maximum obtained recognition rate by using words was 73%. For the digit recognition system, a recognition rate of 98.57% was obtained. PLP and RASTA_PLP showed better performance in the developed pathology assessment systems.

Analysis of Learning Rate Using CPN Algorithm for Hand Written Character Recognition Application

This paper presents the analysis of the learning rate using the Counter Propagation network (CPN) algorithm for the hand written characters recognition application. The recognition process uses the forward only CPN algorithm to recognize the hand written characters. The experimental results obtained with different learning rate values shows that learning rate has large effect on the recognition process. Upper-case English alphabets for a number of different styles gathered from different peoples are used in the analysis for the performance of the CPN algorithm. The obtained recognition rates were 60% to 98% using the CPN for different learning rate value. The experimental results are very encouraging and satisfactory.

A technique to overcome the problem of small size database for automatic speaker recognition

Modeling a system by statistical methods needs large amount of data to train the system. In real life, such data are sometimes not available or hard to collect. Modeling the system with small size database will produce a system with poor performance. In this paper, a method for increasing the size of a speech database is proposed. The method works by generating new samples from the original samples, using combinations of the following methods: speech lengthening, noise addition, and word reversal. To make a proof of concept, a severe test condition is used, in which the original database consists of one sample per speaker, for a speaker recognition system. The system is tested using original samples and the highest obtained recognition rate for mixed genders is 91.41% and that of 93.24% for male only speakers. Key words: Lengthening of samples, noise addition, word reversal, speaker recognition.

Selective Facial Expression Recognition Using fastICA

This paper proposes a facial expression recognition approach based on the combination of fastICA method and neural network classifiers. First we get some special facial expression regions, including eyebrows, eyes and mouth, in which wavelet transform is done to reduce the dimension. Then the fastICA method is used to extract these three facial features. Finally, BP neural network classifier is adopted to recognize facial expression. Experimental on the JAFFE database results show that the method is effective for both dimension reduction and recognition performance in comparison with traditional PCA and ICA method. We have obtained recognition rates as high as 93.33% in categorizing the facial expressions neutral, anger, or sadness. The best average recognition rate achieves 90.48%.

Farsi Font Face Recognition in Letter Level

Most of the presented papers about font recognition, are in block or line level. In this paper a new approach is presented which is able to recognize font of a Farsi document image in letter level. In this approach using the Euclidean distance between spatial descriptors and gradient value in each boundary point of some special Farsi letters in a document image its font is recognized. To implement and evaluate this approach we constructed a dataset consisting of some templates in 25 widely used Farsi fonts and another dataset including 500 Farsi document images. Obtained recognition rate was 98.7%.

DFAspike: A new computational proposition for efficient recognition of epileptic spike in EEG

An automated method has been presented for the detection of epileptic spikes in the electroencephalogram (EEG) using a deterministic finite automata (DFA) and has been named as DFAspike. EEG data (sampled, 256 Hz) files are the inputs to the DFAspike. The DFAspike was tested with different data files containing epileptic spikes. The obtained recognition rate of epileptic spike was 99.13% on an average. This system does not require any kind of prior training or human intrusion. The result shows that the designed system can be very effectively used for the detection of spikes present in the recorded EEG signals.

A Method of Optimal Sensor Decision for Odor Recognition

ABSTRACT In this paper, we propose method of correlation coefficients between sensors by statistical analysis that selects optimal sensors in odor recognition system of selective multi-sensors. The proposed sensor decision method obtains odor data from Metal Oxide Semiconductor(MOS) sensor array and then, we decide optimal sensors based on correlation of obtained odors. First of all, we select total number of 16 sensors eliminated sensor of low response and low reaction rate response among similar sensors. We make up DB using 16 sensors from input odor and we select sensor of low correlation after calculated correlation coefficient of each sensor. Selected sensors eliminate similar sensors' response therefore proposed method are able to decide optimal sensors. We applied to floral scent recognition for performance evaluation of proposed sensors decision method. As a result, application of proposed method with floral scent recognition using correlation coefficient obtained recognition rate of 95.67% case of using 16 sensors while applied floral scent recognition system of proposed sensor decision method confirmed recognition rate of 94.67% using six sensors and 96% using only 8 sensors.Keywords:Sensors decision method, Correlation Coefficient, Floral scent Recognition

냄새 인식을 위한 최적의 센서 결정 방법

ABSTRACT In this paper, we propose method of correlation coefficients between sensors by statistical analysis that selects optimal sensors in odor recognition system of selective multi-sensors. The proposed sensor decision method obtains odor data from Metal Oxide Semiconductor(MOS) sensor array and then, we decide optimal sensors based on correlation of obtained odors. First of all, we select total number of 16 sensors eliminated sensor of low response and low reaction rate response among similar sensors. We make up DB using 16 sensors from input odor and we select sensor of low correlation after calculated correlation coefficient of each sensor. Selected sensors eliminate similar sensors' response therefore proposed method are able to decide optimal sensors. We applied to floral scent recognition for performance evaluation of proposed sensors decision method. As a result, application of proposed method with floral scent recognition using correlation coefficient obtained recognition rate of 95.67% case of using 16 sensors while applied floral scent recognition system of proposed sensor decision method confirmed recognition rate of 94.67% using six sensors and 96% using only 8 sensors.Keywords:Sensors decision method, Correlation Coefficient, Floral scent Recognition