Template Learning Research Articles

Learning templates for artistic portrait lighting analysis.

Lighting is a key factor in creating impressive artistic portraits. In this paper, we propose to analyze portrait lighting by learning templates of lighting styles. Inspired by the experience of artists, we first define several novel features that describe the local contrasts in various face regions. The most informative features are then selected with a stepwise feature pursuit algorithm to derive the templates of various lighting styles. After that, the matching scores that measure the similarity between a testing portrait and those templates are calculated for lighting style classification. Furthermore, we train a regression model by the subjective scores and the feature responses of a template to predict the score of a portrait lighting quality. Based on the templates, a novel face illumination descriptor is defined to measure the difference between two portrait lightings. Experimental results show that the learned templates can well describe the lighting styles, whereas the proposed approach can assess the lighting quality of artistic portraits as human being does.

Circuit Breaker Mechanical Characteristic Parameters Measurement Based on Machine Vision

Aim at the disadvantages of traditional circuit breaker mechanical characteristic parameters test. Get the motion pictures of insulation connecting rod through high-speed camera, using the finite difference method to quickly screen out the motion pictures, and selecting punctuation area as a template for learning, using non-uniform sampling have a template matching, obtain the center coordinates of matching results, time interval is known every frame. Through coordinate changes over time we can obtain mechanical parameters of the circuit breaker accurately, fast, conveniently. Lab VIEW programs achieve the above process automatically.

HIGH-SPEED RAILWAY BRIDGES: CHALLENGES FOR ENGINEERS OF THE CIS

Purpose. National regulatory framework for the CIS does not contain a set of rules sufficient to design highspeed rail lines and related infrastructure. For example, the basic normative document SNiP «Bridges and pipes» there is no «dedicated passenger line» do not take into account the speed of trains, and a variety of dynamic effectsare taken into account empirical coefficients. Methodology. The main features of the system of higher technical education in the CIS are «typing» of course projects and papers, template learning process, the inability to select and change disciplines, departments and teachers. A serious challenge for the student acts as a permanent reduction in the number of classroom hours, with a clear shift towards self-education. At the same time, modern engineering essentially computerized, thanks to the active use of various computer-aided design. Findings. Recommendations for designers of high-speed railway bridges: in most cases sufficient to restrict rod finite elements; modal analysis will make it possible to identify «weak» construction sites, to determine the periods and frequency of free oscillations. Dynamic calculation of time will allow to analyze the stress-strain state of the bridge while driving a train at apredetermined rate, calculate the speed, handling, acceleration. Originality.Today in the CIS appear first private research centers. For example, the Ukrainian Research Laboratory «Dynamics & Structures, Lab.» (www.dystlab.com) was established to study and solve the problems of the dynamics of structures, buildings and structures, as well as related problems in the mechanics of deformable systems. Practical value. In the laboratory «Dynamics & Structures, Lab.» engaged in the development of specialized software – parametric modeling and engineering analysis environment Belinda Structure. Program based on the finite element method, and the simulation is carried out using scripting technologies. The interface of the program includes advanced tools for threedimensional visualization.

Analysis of Contradictions in Online Collaborative Learning using Activity Theory as Analytical Framework

While previous studies have cited the benefits of incorporating online collaborative learning (OCL) in teaching and learning, there are also shortcomings of OCL that should not be ignored. This study aimed to investigate the constraints of incorporating OCL in a Malaysian tertiary classroom using Activity Theory as analytical framework. Activity Theory proved to be useful, particularly for OCL, because it provided a structure for identifying internal contradictions, also referred as tensions or constraints as result of interaction by the components of the OCL. Hence, the objectives of this paper are twofold: to identify constraints of OCL as perceived by students; and to investigate their suggestions or recommendations for improvement of OCL. The findings indicated two keys constraints: technology-related contradictions, which are related to desire for synchronous feedback in forum discussions, cut and paste and plagiarism of ideas, and other technological distractions; and group discussion contradictions. These refer to repetitive and mixed-up postings, clashes on topics of discussion, and discussions being too formal. Suggestions for improvement are reported by students regarding aspects of personalizing an online collaborative learning template; and providing additional support for collaborating online.

Management learning, performance and reward: theory and practice revisited

Purpose – The purpose of this paper is to explore the extent to which organizational learning is recognized through performance management systems as contributing to organizational effectiveness and competitive advantage. Design/methodology/approach – It reviews several pieces of research, employing a wide range of methods, including: content analysis of managers’ reflections; questionnaires completed by managers and mentors; a large-scale survey involving ethnography, interviews and questionnaires; and analysis of documents from professional bodies and management delivery centres. Findings – Genuine integration of individual and organizational goals or transfer of learning from the individual to the organization is not evident. Few qualitative measures of organizational performance are employed. The impact of metrics such as IIP or EFQM on organizational effectiveness is nor discernible. Management learning and development is rarely measured even when it is encouraged by the organization. There is a clear divide between research, teaching and learning and workplace practice. Performance management systems create perceptions of unreliability and inequity. Research limitations/implications – Espousing the value of learning and learning to learn, measuring them accurately and rewarding them with meaningful changes to working life can only improve organizational effectiveness. Research into the few organizations that have successfully embraced triple loop learning in their development of managers may offer a template for transformational learning to sustain competitive advantage. Originality/value – Management development processes have been successful in developing individuals but less successful in achieving organizational development. This paper offers new insights into that gap and the omissions in the metrics by which performance is measured.

Using performance templates for manager and employee learning – in-action

Purpose – This article demonstrates how the performance improvement process, performance templates, may be used for both manager and employee learning. The paper explains the concept, offers theoretical support for it, gives evidence of the efficacy of the process, and explains, in detail, how the process may be applied. Design/methodology/approach – The approach is to present or demonstrate what performance templates are and how a manager and a group of employees work together to create them and use them in the field. The paper presents the approach, explains how it connects with theory, offers empirical support for performance templates, and gives a sequence of events as to their creation and use. Findings – The findings indicate that, in general, the learning and performance approach, performance templates offer an influential and practical tool for both manager and employee learning, and consistent performance improvement. The approach has many applications, although recent research relies mostly on its use in sales organizations. Research limitations/implications – Primarily, the performance template approach has been used with small samples of learners. In the extant studies of the approach, small sample sizes limit the value of statistical analysis. Practical implications – The approach has not been tested with large employee groups (more than 25). The method is labor-intensive and time consuming. Large groups of participants would compromise the use of the approach as individual employee participation could be limited. Originality/value – Performance templates add to the arsenal of tools and methods available to trainers and managers. There is some empirical support for the approach. Performance templates represent both a learning approach and a performance improvement approach as managers and their employees work jointly to create learning and change.

Genetic and brain atrophy markers associated with different psychological phenotypes within ADNI

room. Patients with Alzheimer’s disease (AD) typically have difficulty with tasks requiring ASA and anatomical regions implicated in ASA in the healthy brain are vulnerable in AD, suggesting that ASAmay provide a window on AD pathophysiology. Methods: Here we studied the brain mechanism of the ’cocktail party effect’ using 3T fMRI in a cohort of patients fulfilling consensus criteria for typical AD in relation to age-matched healthy subjects.We designed a ’sparse’ image acquisition, passive listening paradigm to minimise extraneous cognitive or task-related attentional demands, control low-level auditory perceptual features and optimise signal detection. We reasoned that the cocktail party effect would be captured by the differential ASA demands during processing of sound conditions where the auditory target was a learned template (one’s own name) versus a comparably acoustically complex but unfamiliar sound object (one’s own name spectrally rotated). Subjects heard four stimulus conditions: i) their own name interleaved with babble, ii) own name superimposed on babble, iii) own name spectrally rotated (rendered unrecognisable) and interleaved with babble, or iv) own name spectrally rotated and superimposed on babble. Results: Compared with healthy older controls, AD patients showed heightened activation in posterior hippocampus during ’cocktail party’ processing. Conclusions: Our findings suggest that ASA is a model of generic cortical information processing that can be used to probe obligatory pathophysological mechanisms of AD, including abnormally heightened or compensatory processes. This work goes beyond the traditional emphasis on higher cognitive and mnestic functions in AD: besides illuminating an important but poorly understood class of perceptual symptoms, such work promises novel insights into neural network dysfunction underpinning AD.

Efficient multitemplate learning for structured prediction.

Conditional random fields (CRF) and structural support vector machines (structural SVM) are two state-of-the-art methods for structured prediction that captures the interdependencies among output variables. The success of these methods is attributed to the fact that their discriminative models are able to account for overlapping features on all input observations. These features are usually generated by applying a given set of templates on labeled data, but improper templates may lead to degraded performance. To alleviate this issue, in this paper we propose a novel multiple template learning paradigm to learn structured prediction and the importance of each template simultaneously, so that hundreds of arbitrary templates could be added into the learning model without caution. This paradigm can be formulated as a special multiple kernel learning problem with an exponential number of constraints. Then we introduce an efficient cutting-plane algorithm to solve this problem in the primal and present its convergence. We also evaluate the proposed learning paradigm on two widely studied structured prediction tasks, i.e., sequence labeling and dependency parsing. Extensive experimental results show that the proposed method outperforms CRFs and structural SVMs because of exploiting the importance of each template. Complexity analysis and empirical results also show that the proposed method is more efficient than Online multikernel learning on very sparse and high-dimensional data. We further extend this paradigm for structured prediction using generalized p-block norm regularization with p >; 1, and experiments show competitive performances when p ∈ [1,2).

Learning Hybrid Image Templates (HIT) by Information Projection

This paper presents a novel framework for learning a generative image representation-the hybrid image template (HIT) from a small number (i.e., 3 \sim 20) of image examples. Each learned template is composed of, typically, 50 \sim 500 image patches whose geometric attributes (location, scale, orientation) may adapt in a local neighborhood for deformation, and whose appearances are characterized, respectively, by four types of descriptors: local sketch (edge or bar), texture gradients with orientations, flatness regions, and colors. These heterogeneous patches are automatically ranked and selected from a large pool according to their information gains using an information projection framework. Intuitively, a patch has a higher information gain if 1) its feature statistics are consistent within the training examples and are distinctive from the statistics of negative examples (i.e., generic images or examples from other categories); and 2) its feature statistics have less intraclass variations. The learning process pursues the most informative (for either generative or discriminative purpose) patches one at a time and stops when the information gain is within statistical fluctuation. The template is associated with a well-normalized probability model that integrates the heterogeneous feature statistics. This automated feature selection procedure allows our algorithm to scale up to a wide range of image categories, from those with regular shapes to those with stochastic texture. The learned representation captures the intrinsic characteristics of the object or scene categories. We evaluate the hybrid image templates on several public benchmarks, and demonstrate classification performances on par with state-of-the-art methods like HoG+SVM, and when small training sample sizes are used, the proposed system shows a clear advantage.

Egg Recognition in Japanese Quail

Egg recognition allows birds to reduce the costs of heterospecific or conspecific nest parasitism, by allowing them to reject foreign eggs or abandon parasitised clutches. However, the precise phenotypic characteristics of eggs that are recognised and compared are not well understood. Using Japanese quail ( Coturnix japonica), which exhibit extremely high between-female and low within-female variation in egg colour, maculation patterns and shape, I show that females can learn to recognise the salient characteristics of eggs from a given clutch (not necessarily their own) and can use this learned template to discriminate against foreign eggs with a high probability when the eggs are phenotypically distinct. Specifically, female quail appear to use both maculation pattern and, to a lesser extent, egg shape to make these decisions. These findings are discussed with respect to cognitive mechanisms of egg recognition.

Turn Them Loose as Web Detectives on Cold Cases: A Digital Age Remedy for Casebook Boredom and Relevancy

What is it about casebooks that is so challenging for novices and so shunned by upper level students? Do casebooks present such barriers to learning that their knowledge-gaining features remain unexploited by many students? These questions were the proximate cause for Brown’s creation of a learning template that demonstrates how case study does not have to be boringly irrelevant to future professionals. Learning from casebooks, a challengingly foreign experience compared to college, appears to be irrelevant to what lawyers do in practice. Then there is the educationally numbing effect continuous case studies have on law students’ enthusiasm. Brown applies his research and experience by demonstrating a methodological strategy for stimulating students to web investigate their decontextualized “cold” appellate cases to resurrect its human aspects. Initially, this methodology can be employed by 1Ls to bridge the difficulties of learning from cases. It guides novices by suggesting appropriate questions to facilitate better Socratic dialogues. In progressive stages, students are assigned to explore internet “surfing” to re-assemble the case’s context with location maps, party history, judicial profiles, etc. It asks students to use their experiences/imaginations to create a virtual story about how the human or business conflict developed prior to trial. It then uses that virtual recreation to explore lawyering functions by in-class role plays of interviewing, drafting, framing a cause of action; negotiating, and mediating.A Web Detectives Template is offered as a practical, easily applied educator’s process tool to aid novices study skills; to indicate an organizational structure for filing notes, rules, etc; and to suggest questions for those not experienced in asking relevant ones. The Template shows where to integrate horn-book materials, critique judicial analyses, and experience a simulation process of client/attorney interactions that reveals how law and problem solving are intertwined. At its best, this builds long term memory about the law and promotes learning from formerly boring cases.

Learning Active Basis Models by EM-Type Algorithms

EM algorithm is a convenient tool for maximum likelihood model fitting when the data are incomplete or when there are latent variables or hidden states. In this review article we explain that EM algorithm is a natural computational scheme for learning image templates of object categories where the learning is not fully supervised. We represent an image template by an active basis model, which is a linear composition of a selected set of localized, elongated and oriented wavelet elements that are allowed to slightly perturb their locations and orientations to account for the deformations of object shapes. The model can be easily learned when the objects in the training images are of the same pose, and appear at the same location and scale. This is often called supervised learning. In the situation where the objects may appear at different unknown locations, orientations and scales in the training images, we have to incorporate the unknown locations, orientations and scales as latent variables into the image generation process, and learn the template by EM-type algorithms. The E-step imputes the unknown locations, orientations and scales based on the currently learned template. This step can be considered self-supervision, which involves using the current template to recognize the objects in the training images. The M-step then relearns the template based on the imputed locations, orientations and scales, and this is essentially the same as supervised learning. So the EM learning process iterates between recognition and supervised learning. We illustrate this scheme by several experiments.

Matching Objects in Multi-Camera Surveillance without Geometric Constraints

Intelligent visual surveillance by multiple cameras has becoming more and more important in recent years. A novel and effective method to match objects between multiple cameras is proposed in this work, based on the proposed wavelet salient features. Firstly, the background and foreground is separated based on Gaussian Mixture Model (GMM). To achieve a reliable matching objects between cameras, the wavelet salient features consist of both the color feature components and the salient value that are extracted from salient regions, rather than the color feature from the whole object as in the traditional methods. To make the matching more robust against occlusion, we design dynamic learning templates with wavelet salient features which are constructed and updated in one camera, and used to match objects in another camera without any geometry constraints. Experimental results demonstrate the effectiveness of the matching in different views, even when occluded.

Prior Experience Alters the Behavioral Response of Prey to a Nonnative Predator

Anuran larvae (tadpoles) can alter their behavior and morphology in response to predators with which they have coevolved. Furthermore, tadpoles of a few species are capable of learning, which can elicit or reinforce predator-avoidance behaviors. However, it remains unclear how widespread this capacity for learning is among anurans and whether it is biased in favor of evolutionarily familiar predators. Here, we test whether prior experiences will modify the behavioral response of Lowland Leopard Frog Lithobates (Rana) yavapaiensis tadpoles to Green Sunfish Lepomis cyanellus, a recently introduced predator. We exposed focal tadpoles for 10 days to the chemical and visual cues of one of three conditioning treatments: a cricket-fed Green Sunfish, a tadpole-fed Green Sunfish, or a control tank without predator. Subsequently, we measured the swimming activity of focal tadpoles in response to a neutral cue (water) and the chemical cues of Green Sunfish. No difference between conditioning treatments was observed in response to the water cue. In contrast, tadpoles that had previously experienced either of the sunfish conditioning treatments displayed significantly higher swimming activity than control tadpoles for 2-4 min after exposure to the sunfish chemical cues. Our results indicate that the behavior of tadpoles can be altered by prior experiences, even in the absence of alarm cues. In addition to providing another example of learning in tadpoles, our results suggest that tadpoles may have a broad learning template that can be applied to organisms with which they have recently come into contact.

Realtime Kernel based Machine Learning Template Matching (KMLT)

This paper deals with a new approach for the problemof realtime planar templatematching. We considertracking as the estimation of a parametric function between observations and motion and we propose anextension of the learning based approach presented simultaneously by Cootes and al. and by Jurie andDhome to non linear regression functions. The estimation of the linear parameters associated to the basisfunctions (kernel functions) of the model is then achieved using a training set of motions and associatedobservations. We show that the resulting method outperforms the robustness of the linear tracker againstnoisy observations.

Sustained Experiential Learning: Modified Monasticism and Pilgrimage

AbstractThis article outlines a template for sustained experiential learning designed to provide a context for learning the affective and performative as well as intellectual power of religion. This approach was developed for a traditional academic framework, adapting pedagogies developed for experiential learning, aesthetic training, and study abroad, and draws on personal experiences of teaching East Asian religions. The approach integrates intellectual learning with out of class experience to stimulate and enrich the highly personal and often significant questions that may arise upon studying religion and encountering religious practices both in and out of the classroom.

Extracting and Using Translation Templates in an Example-Based Machine Translation System

A bidirectional English-Filipino machine translation system is developed that extracts translation templates and chunks from a given bilingual English-Filipino corpus. These templates and chunks are then used to translate an input English document to Filipino and vice versa. The system extended the similarity and difference translation template learning algorithms of Cicekli and Guvenir (2003) by refining existing templates and deriving templates from previously learned chunks. Chunk alignment, splitting algorithms, and chunk refinement are also introduced in the training process. Correct extraction of similarity templates and chunks during the learning process led to translation with a low word error rate of 15% for a test document whose sentences exactly match the training set, to a high 86% when the test document is different from the training corpus. Using difference templates alone, the resulting translation has a word error rate of 49% to 85%. Combined use of similarity and difference templates resulted in a low word error rate of 18% when the test document contains sentence patterns matching the training set, to a high 85% when the test document is different from the training corpus. Tests also showed that the translation with the highest score selected from a set of candidate translations is consistently the best choice as validated against automatic evaluation methods.

The development of an evidence-based educational framework to facilitate the training of competent rheumatologist ultrasonographers

Musculoskeletal ultrasonography (MSKUS) has been described by some rheumatologists as the 'stethoscope of the joint'. Such enthusiasm is supported by evidence confirming validity and clinical utility in evaluation of musculoskeletal diseases. But if rheumatologist-performed MSKUS is to emulate the impact of cardiologist-performed echocardiography, a number of educational challenges need be addressed. Evaluation of current training reveals the absence of a unified educational structure, ad hoc teaching and assessment and published data are insufficient to make practice and training recommendations specific to rheumatology. Informed by developments in adult-learning theory, we have utilized a competency-based approach to develop an educational programme for rheumatologist ultrasonographers. Fundamental to this process has been accurate functional analysis of the role of the rheumatologist ultrasonographer and precise translation of these data into educational outcomes. This involved a thorough, transparent, iterative, curriculum-defining approach, employing quantitative and qualitative research techniques including interview, questionnaire, Delphi and standard setting methodologies. All relevant stakeholders were engaged, including international US experts and clinical rheumatologists. Outcomes include clarification of the role of a rheumatologist ultrasonographer; definition of appropriate knowledge and skills; establishment of competency standards; a balanced, clearly defined, clinically relevant educational outcome blueprint. Teaching and assessment approaches have been piloted as part of an accessible modular curriculum strategy. Thorough validation and evaluation confirms effectiveness, efficiency and suitability. A comprehensive evidence-based, expert consensus-defined educational framework is proposed that provides a template for teaching and learning and standards for competency assessment. This should facilitate common principles of training, uniform professional practice and a justifiable governance structure.

A Recurrent Fuzzy Coupled Cellular Neural Network System With Automatic Structure and Template Learning

The cellular neural network (CNN) is a powerful technique to mimic the local function of biological neural circuits, especially the human visual pathway system, for real-time image and video processing. Recently, many studies show that an integrated CNN system can solve more complex high-level intelligent problems. In this brief, we extend our previously proposed multi-CNN integrated system, called recurrent fuzzy CNN (RFCNN) which considers uncoupled CNNs only, to automatically learn the proper network structure and parameters simultaneously of coupled CNNs, which is called recurrent fuzzy coupled CNN (RFCCNN). The proposed RFCCNN provides a solution to the current dilemma on the decision of templates and/or fuzzy rules in the existing integrated (fuzzy) CNN systems. For comparison, the capability of the proposed RFCCNN is demonstrated on the same defect inspection problems. Simulation results show that the proposed RFCCNN outperforms the RFCNN

An Exercise in Psychological Growth*

The author presents a case report followed by a self- assessment exercise which was generated spontaneously in collaboration with the patient herself during the treatment process. The lists of questions, designed to provoke thought and stimulate maturation, were found to be quite helpful especially in making important life decisions. It is proposed as a learning tool and template for use by those confronting similar issues.