Search results

The purpose of this paper is to develop an automated human body measurement extraction system using simple inexpensive equipment with minimum requirement of human assistance. This research further leads to the comparison of extracted measurements to established methods to analyze the error. The extracted measurements can be used to assist the production of custom-fit apparel. This is an effort to reduce the cost of expensive 3-D body scanners and to make the system available to the user at home.

A single camera body measurement system is proposed, implemented, and pilot tested. This system involves a personal computer and a webcam operating within a space of controlled lighting. The system will take two images of the user, extract body silhouettes, and perform measurement extraction. The camera is automatically calibrated using the software each time of scanning considering the scanning space. The user will select a front view and a side view among the images captured, and specify the height. In this pilot study, 31 subjects were recruited and the accuracy of 8 human body measurements were compared with the manual measurements and measurements extracted from a commercial 3-D body scanner.

The system achieved reasonable measurement performance within 10 percent accuracy for seven out of the eight measurements, while four out of eight parameters obtained a performance similar to the commercial scanner. It is proved that human body measurement extraction can be done using inexpensive equipment to obtain reasonable results.

This study is aimed at developing a proof-of-concept for inexpensive body scanning system, with an effort to benchmark measurement accuracy, available to an average user providing the ability to acquire self-body measurements to be used to purchase custom-fit apparel. This system can potentially boost the customization of apparel and revolutionize online shopping of custom-fit apparel.

This study focused on how to quantify the similarities of body shape based on the front and side images, and a shape comprehensive index (I_SC) of female upper body shape based on 2D images was proposed.

In total, 190 young women were shot for front and side images, and 18 shape parameters were automatically extracted, including seven angles and 11 ratio parameters. The coefficient of variation method was used to assign different weights for related parameters, and the I_SC was calculated to describe the body shape of each subject. Five cross-sectional curves of the upper body (e.g. shoulder, chest, waist, abdomen and hip) were selected for exploring the range of shape similarity.

According to the value of I_SC, if the difference among the subjects is within the range of ±0.02, their body shapes can be regarded as similar, and the subject with the minimum distance is considered as the most similar. Error results show that the error range of the angle parameter is from 0.2° to 3.6° and the ratio range is from 0.001 to 0.119. Moreover, the t-test value among the parameters of the similar body is above 0.05, indicating that there is no significant difference for the upper body shape of the similar groups.

This method can quantify body shapes with the upper body characteristics of young women instead of subjective judgment. The study can be extended to other parts of the body and can also provide a new thought for shape similarity retrieval based on 2D images.

The paper aims to introduce an intelligent recognition system for viewpoint variations of gait and speech. It proposes a convolutional neural network-based capsule network (CNN-CapsNet) model and outlining the performance of the system in recognition of gait and speech variations. The proposed intelligent system mainly focuses on relative spatial hierarchies between gait features in the entities of the image due to translational invariances in sub-sampling and speech variations.

This proposed work CNN-CapsNet is mainly used for automatic learning of feature representations based on CNN and used capsule vectors as neurons to encode all the spatial information of an image by adapting equal variances to change in viewpoint. The proposed study will resolve the discrepancies caused by cofactors and gait recognition between opinions based on a model of CNN-CapsNet.

This research work provides recognition of signal, biometric-based gait recognition and sound/speech analysis. Empirical evaluations are conducted on three aspects of scenarios, namely fixed-view, cross-view and multi-view conditions. The main parameters for recognition of gait are speed, change in clothes, subjects walking with carrying object and intensity of light.

The proposed CNN-CapsNet has some limitations when considering for detecting the walking targets from surveillance videos considering multimodal fusion approaches using hardware sensor devices. It can also act as a pre-requisite tool to analyze, identify, detect and verify the malware practices.

This research work includes for detecting the walking targets from surveillance videos considering multimodal fusion approaches using hardware sensor devices. It can also act as a pre-requisite tool to analyze, identify, detect and verify the malware practices.

This proposed research work proves to be performing better for the recognition of gait and speech when compared with other techniques.

With the rapid advancement of computer information technology, the traditional clothing industry has stridden towards automation and digitization that drive the growth of electronic commerce and line retailing. The purpose of this paper is to propose an approach on 3D upper body modelling based on the body measurements extracted by non-contact anthropometry.

Based on the frontal and side images of the human body, the body sizes were extracted through silhouette extraction, identification of landmarks and girth prediction. The generation rules of 15 characteristic cross-sectional curves were established using a method “feature points – inserted points – feature curves – basic surface – mannequin”. The feature points of each position were determined at each curve, such as the side neck point, front neck point, shoulder point, bust point, and bust root point and so on to get the cross-sections, and then some feature points were inserted at the curves according to the widths and depths to establish the calculative models. For example, there are 18 points distributed at the bust cross-sectional curve to determine the shape.

The final mannequin could describe the basic characteristics of a human body, and the shape of the feature curves could also fit the body type to provide basis for the future research on automatic pattern generation.

This study can realize the 3D virtual modelling of female upper body and the automatic generation of the individualized apparel patterns based on the frontal and side images.

The purpose of this paper is to present a novel multimodal approach to the problem of planning and performing a reliable grasping action on unmodeled objects.

The robotic system is composed of three main components. The first is a conceptual manipulation framework based on grasping primitives. The second component is a visual processing module that uses stereo images and biologically inspired algorithms to accurately estimate pose, size, and shape of an unmodeled target object. A grasp action is planned and executed by the third component of the system, a reactive controller that uses tactile feedback to compensate possible inaccuracies and thus complete the grasp even in difficult or unexpected conditions.

Theoretical analysis and experimental results have shown that the proposed approach to grasping based on the concurrent use of complementary sensory modalities, is very promising and suitable even for changing, dynamic environments.

Additional setups with more complicate shapes are being investigated, and each module is being improved both in hardware and software.

This paper introduces a novel, robust, and flexible grasping system based on multimodal integration.

The current studies on the clothing silhouette are very limited. The purpose of this paper is to propose an innovative framework to intelligently identify the womenswear silhouette with the latest computer technologies. To clearly define the womenswear silhouette, an accurate numerical definition is proposed.

The study first processes and segments the useful parts on the static catwalk image data files following existing graphic extraction approaches. Six basic alphabetical womenswear silhouette types are selected and numerically defined. Then, the proposed framework automatically classifies the six basic womenswear silhouettes considering the different slopes between three main clothing parts. Six clothing situations are discovered according to different designs and the detailed cases are systematically categorized. In addition, aspects influencing the judgment of the clothing silhouettes such as the skin, the background, the drastic change points are also considered. The proposed silhouette definition and identification framework is novel and proved accurate.

The proposed definition and identification framework of womenswear silhouettes have been proved a viable approach that is fully compatible with the current computer technologies. The validation study shows that the presented identification procedure has a desirable accuracy over 90 percent.

The proposed methodology develops brand new standards to numerically define and identify the womenswear silhouette, which was not available in the past. Besides, the measurement, the identification and the classification procedures are fully validated by the image data collected from 14 world famous brands over 11 consecutive seasons. It is shown that the proposed numerical framework of the womenswear silhouettes is a robust one, considering all of the observed design variabilities.

The purpose of this paper is to provide algorithms of the automatic landmark extraction software program that are applicable for any torso shape.

In this study, Automatic Landmark Identification (AULID), an automatic landmark extraction software program, was developed to extract consistent landmark locations from any torso shape. A methodology of geometrical characteristics of the body surfaces around each landmark was used for the algorithms and implemented with C++. The accuracy of the AULID was tested on various torso shapes. The verification methodology consisted of mean difference (MD), mean absolute differences (MAD), and one‐way analysis of variance. Duncan test for multiple comparisons was used to evaluate the significant differences of MAD values among different torso groups. The MAD values were compared to the anthropometric survey allowable errors.

The algorithms of AULID provided both accuracy and consistency of identifying landmarks on any body torso types.

Most 3D body scanning systems often show landmark location errors when dealing with nonstandard body shapes. None of automatic landmark extraction software program provides consistency of identifying landmarks in various body shapes. However, algorithms of AULID, an automatic landmark extraction software program, in this study are only consistent definitions for identifying landmarks in any torso shape.

The purpose of this paper is to present a preliminary solution to address the problem of estimating human pose and trajectory by an aerial robot with a monocular camera in near real time.

The distinguishing feature of the solution is a dynamic classifier selection architecture. Each video frame is corrected for perspective using projective transformation. Then, a silhouette is extracted as a Histogram of Oriented Gradients (HOG). The HOG is then classified using a dynamic classifier. A class is defined as a pose-viewpoint pair, and a total of 64 classes are defined to represent a forward walking and turning gait sequence. The dynamic classifier consists of a Support Vector Machine (SVM) classifier C64 that recognizes all 64 classes, and 64 SVM classifiers that recognize four classes each – these four classes are chosen based on the temporal relationship between them, dictated by the gait sequence.

The solution provides three main advantages: first, classification is efficient due to dynamic selection (4-class vs 64-class classification). Second, classification errors are confined to neighbors of the true viewpoints. This means a wrongly estimated viewpoint is at most an adjacent viewpoint of the true viewpoint, enabling fast recovery from incorrect estimations. Third, the robust temporal relationship between poses is used to resolve the left-right ambiguities of human silhouettes.

Experiments conducted on both fronto-parallel videos and aerial videos confirm that the solution can achieve accurate pose and trajectory estimation for these different kinds of videos. For example, the “walking on an 8-shaped path” data set (1,652 frames) can achieve the following estimation accuracies: 85 percent for viewpoints and 98.14 percent for poses.

Document management is growing in importance proportionate to the growth of unstructured data, and its applications are increasing from process benchmarking to customer relationship management and so on. The purpose of this paper is to improve important components of document management that is keyword extraction and document clustering. It is achieved through knowledge extraction by updating the phrase document matrix. The objective is to manage documents by extending the phrase document matrix and achieve refined clusters. The study achieves consistency in cluster quality in spite of the increasing size of data set. Domain independence of the proposed method is tested and compared with other methods.

In this paper, a synset-based phrase document matrix construction method is proposed where semantically similar phrases are grouped to reduce the dimension curse. When a large collection of documents is to be processed, it includes some documents that are very much related to the topic of interest known as model documents and also the documents that deviate from the topic of interest. These non-relevant documents may affect the cluster quality. The first step in knowledge extraction from the unstructured textual data is converting it into structured form either as term frequency-inverse document frequency matrix or as phrase document matrix. Once in structured form, a range of mining algorithms from classification to clustering can be applied.

In the enhanced approach, the model documents are used to extract key phrases with synset groups, whereas the other documents participate in the construction of the feature matrix. It gives a better feature vector representation and improved cluster quality.

Various applications that require managing of unstructured documents can use this approach by specifically incorporating the domain knowledge with a thesaurus.

Experiment pertaining to the academic domain is presented that categorizes research papers according to the context and topic, and this will help academicians to organize and build knowledge in a better way. The grouping and feature extraction for resume data can facilitate the candidate selection process.

Applications like knowledge management, clustering of search engine results, different recommender systems like hotel recommender, task recommender, and so on, will benefit from this study. Hence, the study contributes to improving document management in business domains or areas of interest of its users from various strata’s of society.

The study proposed an improvement to document management approach that can be applied in various domains. The efficacy of the proposed approach and its enhancement is validated on three different data sets of well-articulated documents from data sets such as biography, resume and research papers. These results can be used for benchmarking further work carried out in these areas.

The overall goal of the data mining process is to extract information from an extensive data set and make it understandable for further use. When working with large volumes of unstructured data in research information systems, it is necessary to divide the information into logical groupings after examining their quality before attempting to analyze it. On the other hand, data quality results are valuable resources for defining quality excellence programs of any information system. Hence, the purpose of this study is to discover and extract knowledge to evaluate and improve data quality in research information systems.

Clustering in data analysis and exploiting the outputs allows practitioners to gain an in-depth and extensive look at their information to form some logical structures based on what they have found. In this study, data extracted from an information system are used in the first stage. Then, the data quality results are classified into an organized structure based on data quality dimension standards. Next, clustering algorithms (K-Means), density-based clustering (density-based spatial clustering of applications with noise [DBSCAN]) and hierarchical clustering (balanced iterative reducing and clustering using hierarchies [BIRCH]) are applied to compare and find the most appropriate clustering algorithms in the research information system.

This paper showed that quality control results of an information system could be categorized through well-known data quality dimensions, including precision, accuracy, completeness, consistency, reputation and timeliness. Furthermore, among different well-known clustering approaches, the BIRCH algorithm of hierarchical clustering methods performs better in data clustering and gives the highest silhouette coefficient value. Next in line is the DBSCAN method, which performs better than the K-Means method.

In the data quality assessment process, the discrepancies identified and the lack of proper classification for inconsistent data have led to unstructured reports, making the statistical analysis of qualitative metadata problems difficult and thus impossible to root out the observed errors. Therefore, in this study, the evaluation results of data quality have been categorized into various data quality dimensions, based on which multiple analyses have been performed in the form of data mining methods.

Although several pieces of research have been conducted to assess data quality results of research information systems, knowledge extraction from obtained data quality scores is a crucial work that has rarely been studied in the literature. Besides, clustering in data quality analysis and exploiting the outputs allows practitioners to gain an in-depth and extensive look at their information to form some logical structures based on what they have found.