Human Subjective Judgment Research Articles

Utilizing Dictionary Learning and Machine Learning for Blind Quality Assessment of 3-D Images

In recent years, perceptual objective quality assessment of 3-D images has become an intense focus of research. In this paper, we propose an efficient blind 3-D image quality assessment (IQA) metric that utilizes binocular vision-based dictionary learning (DL) and ${k}$ -nearest-neighbors (KNN)-based machine learning (ML) to more accurately align with human subjective judgments. More specifically, in the DL stage, histogram representations from the local patterns of simple and complex cells are concatenated to form basic feature vectors. Then, by using a collaborative representation algorithm, the learned binocular quality-aware features of the distorted 3-D image can be efficiently represented by a linear combination of only a few of these basic feature vectors. In the ML stage, we intuitively simulate the complex high-level behaviors of human perceptual activity with KNN-based ML, which transfers the weighted human subjective quality scores from the annotated 3-D images to the query 3-D image. Our results using three standard subject-rated 3-D-IQA databases confirm that the proposed metric consistently aligns with the subjective ratings and outperforms many representative blind metrics.

A shallow convolutional neural network for blind image sharpness assessment.

Blind image quality assessment can be modeled as feature extraction followed by score prediction. It necessitates considerable expertise and efforts to handcraft features for optimal representation of perceptual image quality. This paper addresses blind image sharpness assessment by using a shallow convolutional neural network (CNN). The network takes single feature layer to unearth intrinsic features for image sharpness representation and utilizes multilayer perceptron (MLP) to rate image quality. Different from traditional methods, CNN integrates feature extraction and score prediction into an optimization procedure and retrieves features automatically from raw images. Moreover, its prediction performance can be enhanced by replacing MLP with general regression neural network (GRNN) and support vector regression (SVR). Experiments on Gaussian blur images from LIVE-II, CSIQ, TID2008 and TID2013 demonstrate that CNN features with SVR achieves the best overall performance, indicating high correlation with human subjective judgment.

Dominance based fuzzy decision support framework for g-resilient (ecosilient) supplier selection: an empirical modelling

A novel decision support framework has been proposed herein to solve supplier selection problems by considering green as well as resiliency criteria, simultaneously. An expert team involving a finite number of Decision-Makers (DMs) is to be employed to obtain degree of importance of each of the evaluation criteria and also to assign an appropriate rating for each supplier alternative with respect to the criteria. Since subjective human judgment often bears some sort of ambiguity as well as vagueness, application of fuzzy set theory can be recommended in this context. The present work represents a dominance based decision support system in the light of fuzzy set theory which is basically a simplified version of TODIM and PROMETHEE. It explores dominance between two alternatives with respect to a particular criterion, based on which a global dominance measure is computed to facilitate ranking order of candidate suppliers. The approach proposed in this research seems straightforward which can exclude complex and tedious computational steps of TODIM as well as PROMETHEE. Application potential of the proposed dominance based fuzzy decision-making approach has been compared to that of Fuzzy-TOPSIS and Fuzzy-VIKOR. Good agreement has been observed. In addition to this, the concept of a unique performance index i.e. ‘g-resilient’ (also called ‘ecosilient’) index has been introduced to help in assessing suppliers’ overall performance extent. The work has also been extended to identify the ill (poor)-performing areas in which suppliers are lagging and seek further improvement towards enhancing g-resilient performance level.

Green supplier selection in fuzzy context: a decision-making scenario on application of fuzzy-MULTIMOORA

Green supply chain management has become an important avenue in current business scenario. The concept of GSCM is to integrate environmental thinking into traditional supply chain management. A firm's sustainability performance is greatly influenced by appropriate supplier selection in the green supply chain context. In selecting appropriate green supplier, various green criteria need to be considered along with traditional supplier selection criteria. The decision making in the context of green supplier selection becomes much more complex due to involvement of subjective selection criteria. Subjective human judgment often bears ambiguity and vagueness in the decision making; whilst fuzzy set theory overcomes the challenges of imprecise and inconsistent human judgment in ambiguous decision environment. In this context, a decision making scenario in relation to green supplier selection has been articulated in this paper aiming to investigate the applicability of MULTIMOORA in fuzzy setting (F-MULTIMOORA). The ranking order of candidate suppliers has been compared to that of fuzzy-TOPSIS. Finally, an attempt has been made to determine a unique quantitative performance index for each of the green suppliers; based on which a ranking order has also been arrived.

RETRACTED ARTICLE: Quality assessment for virtual reality technology based on real scene

Virtual reality technology is a new display technology, which provides users with real viewing experience. As known, most of the virtual reality display through stereoscopic images. However, image quality will be influenced by the collection, storage and transmission process. If the stereoscopic image quality in the virtual reality technology is seriously damaged, the user will feel uncomfortable, and this can even cause healthy problems. In this paper, we establish a set of accurate and effective evaluations for the virtual reality. In the preprocessing, we segment the original reference and distorted image into binocular regions and monocular regions. Then, the Information-weighted SSIM (IW-SSIM) or Information-weighted PSNR (IW-PSNR) values over the monocular regions are applied to obtain the IW-score. At the same time, the Stereo-weighted-SSIM (SW-SSIM) or Stereo-weighted-PSNR (SW-PSNR) can be used to calculate the SW-score. Finally, we pool the stereoscopic images score by combing the IW-score and SW-score. Experiments show that our method is very consistent with human subjective judgment standard in the evaluation of virtual reality technology.

Blind image quality assessment via probabilistic latent semantic analysis.

We propose a blind image quality assessment that is highly unsupervised and training free. The new method is based on the hypothesis that the effect caused by distortion can be expressed by certain latent characteristics. Combined with probabilistic latent semantic analysis, the latent characteristics can be discovered by applying a topic model over a visual word dictionary. Four distortion-affected features are extracted to form the visual words in the dictionary: (1) the block-based local histogram; (2) the block-based local mean value; (3) the mean value of contrast within a block; (4) the variance of contrast within a block. Based on the dictionary, the latent topics in the images can be discovered. The discrepancy between the frequency of the topics in an unfamiliar image and a large number of pristine images is applied to measure the image quality. Experimental results for four open databases show that the newly proposed method correlates well with human subjective judgments of diversely distorted images.

Quality assessment metric of stereo images considering cyclopean integration and visual saliency

In recent years, there has been great progress in the wider use of three-dimensional (3D) technologies. With increasing sources of 3D content, a useful tool is needed to evaluate the perceived quality of the 3D videos/images. This paper puts forward a framework to evaluate the quality of stereoscopic images contaminated by possible symmetric or asymmetric distortions. Human visual system (HVS) studies reveal that binocular combination models and visual saliency are the two key factors for the stereoscopic image quality assessment (SIQA) metric. Therefore inspired by such findings in HVS, this paper proposes a novel saliency map in SIQA metric for the cyclopean image called “cyclopean saliency”, which avoids complex calculations and produces good results in detecting saliency regions. Moreover, experimental results show that our metric significantly outperforms conventional 2D quality metrics and yields higher correlations with human subjective judgment than the state-of-art SIQA metrics. 3D saliency performance is also compared with “cyclopean saliency” in SIQA. It is noticed that the proposed metric is applicable to both symmetric and asymmetric distortions. It can thus be concluded that the proposed SIQA metric can provide an effective evaluation tool to assess stereoscopic image quality.

Binocular Responses for No-Reference 3D Image Quality Assessment

Perceptual quality assessment of distorted three-dimensional (3D) images has become a fundamental yet challenging issue in the field of 3D imaging. In this paper, we propose a general-purpose blind/no-reference (NR) 3D image quality assessment (IQA) metric that utilizes the complementary local patterns (the local magnitude pattern and the proposed generalized local directional pattern) of binocular energy response (BER) and binocular rivalry response (BRR). The main technical contribution of this research is that binocular visual perception and local structural distribution are considered for NR 3D-IQA. More specifically, the metric simulates the binocular visual perception using BER and BRR. Subsequently, the local patterns of the binocular responses’ encoding maps are used to form various binocular quality-predictive features, which will change in the presence of distortions. After feature extraction, we use k -nearest neighbors-based machine learning to drive the overall quality score. We tested our proposed metric against two publicly available 3D databases; these tests confirm that the proposed metric's results consistently align with human subjective judgments.

On Predicting Visual Comfort of Stereoscopic Images: A Learning to Rank Based Approach

Predicting the degree of experienced visual comfort in the context of stereoscopic 3-D (S3D) viewing is particularly challenging. In this letter, a simple yet effective visual comfort assessment (VCA) approach for stereoscopic images is proposed from the perspective of learning to rank (L2R). The proposed L2R-based VCA (L2R-VCA) approach is inspired by the traditional absolute categorical rating (ACR) methodology in subjective study and is to characterize the qualitative description behavior of human subjective study. Experimental results on our recently built database confirm the promising performance of the proposed L2R-VCA approach, yielding higher consistency with human subject judgment results.

An Intelligent Complex Event Processing withDNumbers under Fuzzy Environment

Efficient matching of incoming mass events to persistent queries is fundamental to complex event processing systems. Event matching based on pattern rule is an important feature of complex event processing engine. However, the intrinsic uncertainty in pattern rules which are predecided by experts increases the difficulties of effective complex event processing. It inevitably involves various types of the intrinsic uncertainty, such as imprecision, fuzziness, and incompleteness, due to the inability of human beings subjective judgment. Nevertheless,Dnumbers is a new mathematic tool to model uncertainty, since it ignores the condition that elements on the frame must be mutually exclusive. To address the above issues, an intelligent complex event processing method withDnumbers under fuzzy environment is proposed based on the Technique for Order Preferences by Similarity to an Ideal Solution (TOPSIS) method. The novel method can fully support decision making in complex event processing systems. Finally, a numerical example is provided to evaluate the efficiency of the proposed method.

A Modified TOPSIS Method Based onDNumbers and Its Applications in Human Resources Selection

Multicriteria decision-making (MCDM) is an important branch of operations research which composes multiple-criteria to make decision. TOPSIS is an effective method in handling MCDM problem, while there still exist some shortcomings about it. Upon facing the MCDM problem, various types of uncertainty are inevitable such as incompleteness, fuzziness, and imprecision result from the powerlessness of human beings subjective judgment. However, the TOPSIS method cannot adequately deal with these types of uncertainties. In this paper, aD-TOPSIS method is proposed for MCDM problem based on a new effective and feasible representation of uncertain information, calledDnumbers. TheD-TOPSIS method is an extension of the classical TOPSIS method. Within the proposed method,Dnumbers theory denotes the decision matrix given by experts considering the interrelation of multicriteria. An application about human resources selection, which essentially is a multicriteria decision-making problem, is conducted to demonstrate the effectiveness of the proposedD-TOPSIS method.

Modeling the initial conditions of interacting galaxy pairs using Identikit

We develop and test an automated technique to model the dynamics of interacting galaxy pairs. We use Identikit (Barnes & Hibbard 2009, Barnes 2011) as a tool for modeling and matching the morphology and kinematics of the interacting pairs of equal-mass galaxies. In order to reduce the effect of subjective human judgement, we automate the selection of phase-space regions used to match simulations to data, and we explore how selection of these regions affects the random uncertainties of parameters in the best-fit model. In this work, we use an independent set of GADGET SPH simulations as input data to determine the systematic bias in the measured encounter parameters based on the known initial conditions of these simulations. We test both cold gas and young stellar components in the GADGET simulations to explore the effect of choosing HI vs. H$\alpha$ as the line of sight velocity tracer. We find that we can group the results into tests with good, fair, and poor convergence based on the distribution of parameters of models close to the best-fit model. For tests with good and fair convergence, we rule out large fractions of parameter space and recover merger stage, eccentricity, pericentric distance, viewing angle, and initial disc orientations within 3$\sigma$ of the correct value. All of tests on prograde-prograde systems have either good or fair convergence. The results of tests on edge-on discs are less biased than face-on tests. Retrograde and polar systems do not converge and may require constraints from regions other than the tidal tails and bridges.

Extension of TODIM combined with grey numbers: an integrated decision making module

Purpose – Decision making is the task of selecting the most appropriate alternative among a finite set of possible alternatives with respect to some attributes. The attributes may be subjective or objective (or combination of both), depending upon the situation; requirements may also be conflicting. In practice, most of the real-world decision-making problems are based on subjective evaluation criteria which are basically ill-defined and vague. Since subjective human judgment bears ambiguity and vagueness in the decision making; application of grey numbers set theory may be proved fruitful in this context. The paper aims to discuss these issues. Design/methodology/approach – Owing to the advantages of grey numbers set theory in tackling subjectivity in decision making; the crisp-TODIM needs to be extended by integrating with grey numbers set theory in order to facilitate decision making consisting of subjective data. Hence, the unified objective of this paper is to propose a grey-based TODIM approach in the context of decision making. Findings – Application potential of grey-TODIM has been demonstrated through a case empirical robot selection problem. Result obtained thereof, has also been compared to that of existing grey-based decision support systems available in literature. Originality/value – Application potential of grey-based decision support systems (grey-TOPSIS, grey analysis, grey-MOORA) have been highlighted in available literature resource. However, the shortcoming of these approaches is that they do not consider decision-makers’ risk attitude while decision making. TODIM method is derived from the philosophy of Cumulative Prospect Theory (CPT) which considers risk averting attitude of the decision maker in case of gain and risk seeking attitude in case of loss, while comparing dominance between two alternatives with respect to a particular criterion. Hence, this paper contributes a mathematical foundation of TODIM coupled with grey numbers set theory for logical decision making.

Three-dimensional visual comfort assessment via preference learning

Three-dimensional (3-D) visual comfort assessment (VCA) is a particularly important and challenging topic, which involves automatically predicting the degree of visual comfort in line with human subjective judgment. State-of-the-art VCA models typically focus on minimizing the distance between predicted visual comfort scores and subjective mean opinion scores (MOSs) by training a regression model. However, obtaining precise MOSs is often expensive and time-consuming, which greatly constrains the extension of existing MOS-aware VCA models. This study is inspired by the fact that humans tend to conduct a preference judgment between two stereoscopic images in terms of visual comfort. We propose to train a robust VCA model on a set of preference labels instead of MOSs. The preference label, representing the relative visual comfort of preference stereoscopic image pairs (PSIPs), is generally precise and can be obtained at much lower cost compared with MOS. More specifically, some representative stereoscopic images are first selected to generate the PSIP training set. Then, we use a support vector machine to learn a preference classification model by taking a differential feature vector and the corresponding preference label of each PSIP as input. Finally, given a testing sample, by considering a full-round paired comparison with all the selected representative stereoscopic images, the visual comfort score can be estimated via a simple linear mapping strategy. Experimental results on our newly built 3-D image database demonstrate that the proposed method can achieve a better performance compared with the models trained on MOSs.

An automatic visual inspection system for cone surface defects

Limited to the impact of human visual sensitivity and individual subjective judgments, traditionally manual detection method is unable to accurately and reliably capture the cone surface defects, and will easily cause a lot of missing and false detection. In this paper, we propose a novel machine vision system for cone surface defect inspection. Specifically, we apply gradient and regional consistency detection algorithms to check the surface defects. And then the detection results will be sent via the serial port to the mechanical separation system, and the defective cones will be peeled off the production line by motor. The proposed system can improve the defect detection efficiency and product quality, also can reduce production costs. Experimental results demonstrate that the proposed approach can obtain superior performance over traditional manual method and has high detection efficiency.

Crack Detection using a Texture Analysis-based Technique for Visual Bridge Inspection

Bridge inspection is a pathway to bridge condition rating assessment, and is an essential element of any bridge management system (BMS). The success of a BMS is highly dependent on the quality of bridge inspection outcomes and accurate estimation of future bridge condition ratings. However, existing visual bridge inspection methods suffer several limitations due to human subjective judgment. In order to minimise such limitations, a feasibility study has been performed to enhance the current visual inspection method using optical image processing techniques. However, the accuracy of the inspection outcomes still requires further improvement. This paper proposes an automatic bridge inspection approach employing wavelet-based image features along with support vector machines (SVM) for automatic detection of cracks in bridge images. A two-stage approach is followed, in the first stage, a decision is made as whether an image should undergo a pre-processing step (depending on image characteristics); in the second stage, wavelet features are extracted from the image using a sliding window texture analysis-based technique. Consequently, an average accuracy of 92% (effect of training image types on accuracy) is obtained even when undertaking experiments with noisy and complex bridge images.

Temporal Video Quality Model Accounting for Variable Frame Delay Distortions

We announce a new video quality model (VQM) that accounts for the perceptual impact of variable frame delays (VFD) in videos with demonstrated top performance on the laboratory for image and video engineering (LIVE) mobile video quality assessment (VQA) database. This model, called VQM_VFD, uses perceptual features extracted from spatial-temporal blocks spanning fixed angular extents and a long edge detection filter. VQM_VFD predicts video quality by measuring multiple frame delays using perception based parameters to track subjective quality over time. In the performance analysis of VQM_VFD, we evaluated its efficacy at predicting human opinions of visual quality. A detailed correlation analysis and statistical hypothesis testing show that VQM_VFD accurately predicts human subjective judgments and substantially outperforms top-performing image quality assessment and VQA models previously tested on the LIVE mobile VQA database. VQM_VFD achieved the best performance on the mobile and tablet studies of the LIVE mobile VQA database for simulated compression, wireless packet-loss, and rate adaptation, but not for temporal dynamics. These results validate the new model and warrant a hard release of the VQM_VFD algorithm. It is freely available for any purpose, commercial, or noncommercial at http://www.its.bldrdoc.gov/vqm/ .

Fuzzy logic based method to measure degree of lean activity in manufacturing industry

Lean manufacturing is gaining popularity as an approach that can achieve significant performance improvement in the industry. However, the application of lean manufacturing is not an easy process. To reach the level of full implementation of lean manufacturing takes a long time and during that time the continuous improvement must be made. In the process of continuous improvement, lean manufacturing assessment is required. One form of assessment is to measure the degree of lean implementation. However, it is the complexity involved in the measure of degree of leanness. This complexity arises due to (a) the inherent multi-dimensional concept of leanness (b) unavailability manufacturing practice database that can be used as a benchmark in assessing the degree of leanness and (c) the necessity for the application of subjective human judgement on lean practices which involve vagueness and bias due to variation of evaluator's knowledge and experience. In this paper a method to deal with the multi-dimensional concept, unavailability benchmark and uncertainty, which arises from the subjective and vague human judgement for the measurement of degree of leanness, is proposed. The multi-dimensional concept involving a variety of components of lean practices is measured in order to arrive at a measure for the lean activity of a given organization. It is constructed from primary and secondary data involving a comprehensive literature review and validated with interviews with a set of sample organizations representing the entire spectrum of the industry. The vagueness of subjective human judgement on degree of application of lean practices is modelled by fuzzy number in conjunction with an additional consideration related to the length of lean practice implementation and the use of multi-evaluators. Value stream mapping is used in scoring the degree of implementation of lean so the use of benchmark is not necessary. Some results from an initial survey from a sample of respondents from the manufacturing industry in Indonesia are presented to illustrate the applicability and potential strength of the proposed method.

Blind image quality assessment using a reciprocal singular value curve

The reciprocal singular value curves of natural images resemble inverse power functions. The bending degree of the reciprocal singular value curve varies with distortion type and severity. We describe two new general blind image quality assessment (IQA) indices that respectively use the area and curvature of image reciprocal singular value curves. These two methods almost require very little prior knowledge of any image or distortion nor any process of training, and they can handle multiple unknown distortions, hence they are no-training methods. Experimental results on five simulated databases show that the proposed algorithms deliver quality predictions that have high correlation with human subjective judgments, and that are competitive with other blind IQA models.

Universal blind image quality assessment using contourlet transform and singular-value decomposition

Most current state-of-the-art blind image quality assessment (IQA) algorithms usually require process training or learning. Here, we have developed a completely blind IQA model that uses features derived from an image’s contourlet transform and singular-value decomposition. The model is used to build algorithms that can predict image quality without any training or any prior knowledge of the images or their distortions. The new method consists of three steps: first, the contourlet transform is used on the image to obtain detailed high-frequency structural information from the image; second, the singular values of the just-obtained “structural image” are computed; and finally, two new universal blind IQA indices are constructed utilizing the area and slope of the truncated singular-value curves of the “structural image.” Experimental results on three open databases show that the proposed algorithms deliver quality predictions that have high correlations against human subjective judgments and are highly competitive with the state-of-the-art.