Actions are spatio-temporal patterns which can be characterized by collections of invariant features. Detection actions is to find the re-occurrences (e.g. through pattern matching) such patterns. This paper addresses two critical issues in matching-based action detection: (1) efficiency search 3D videos and (2) tolerance intra-pattern variations actions. Our contributions two-fold. First, we propose a discriminative matching called naive-Bayes based mutual information maximization (NBMIM)...

Linear mixed-effects models are frequently used to analyze repeated measures data, because they model flexibly the within-subject correlation often present in this type of data. The most popular linear for a continuous response assumes normal distributions random effects and errors, making it sensitive outliers. Such outliers more problematic than fixed-effects models, may occur effects, or both, them harder be detected practice. Motivated by real dataset from an orthodontic study, we...

Hand motion capture is one of the most important parts gesture interfaces. Many current approaches to this task generally involve a formidable nonlinear optimization problem in large search space. Motion can be achieved more cost-efficiently when considering constraints hand. Although some represented as equalities or inequalities, there exist many which cannot explicitly represented. In paper, we propose learning approach model hand configuration space directly. The redundancy eliminated by...

Analyzing hand gestures is a comprehensive task involving motion modeling, analysis, pattern recognition, machine learning and even psycholinguistic studies. A review of various techniques in recognition needed. Due to the multidisciplinary nature this research topic, we cannot include all works literature. Rather than function as thorough paper, article serves tutorial topic. We study 3-D models, articulated analysis methods, gesture employed current research. conclude with some thoughts...

Since the human hand is highly articulated and deformable, posture recognition a challenging example in research on view-independent object recognition. Due to difficulties of model-based approach, appearance-based learning approach promising handle large variation visual inputs. However, generalization many proposed supervised methods this problem often suffers from insufficiency labeled training data. This paper describes an alleviate difficulty by adding unlabeled set. Combining...

Given a spectral library, sparse unmixing aims at finding the optimal subset of endmembers from it to model each pixel in hyperspectral scene. However, still remains challenging task due usually high mutual coherence library. In this paper, we exploit priori information image alleviate difficulty. It assumes that some materials library are known exist Such can be obtained via field investigation or data analysis. Then, propose novel incorporate into unmixing. Based on alternating direction...

We introduce the novel Diffusion Visual Programmer (DVP), a neuro-symbolic image translation framework. Our proposed DVP seamlessly embeds condition-flexible diffusion model within GPT architecture, orchestrating coherent sequence of visual programs (i.e., computer vision models) for various pro-symbolic steps, which span RoI identification, style transfer, and position manipulation, facilitating transparent controllable processes. Extensive experiments demonstrate DVP's remarkable...

(1) Purpose: Previous studies investigated the positive relationship between professional identity and career satisfaction in teachers, but underlying reasons were not explored. Therefore, present study explores mediating effects of two variables, namely, psychological empowerment work engagement on satisfaction. (2) Method: The used scale, Utrecht Work Engagement scale to investigate 2104 teachers (Mage = 39.50 years, SD 8.74) a province China. demographic variables (e.g., gender, age,...

A new method for visual tracking of articulated objects is presented. Analyzing motion challenging because the dimensionality increase potentially demands tremendous computation. To ease this problem, we propose an approach that analyzes subparts locally while reinforcing structural constraints at mean time. The computational model proposed based on a dynamic Markov network, generative which characterizes dynamics and image observations each individual subpart as well among different...

Dynamic textures are sequences of images moving scenes that exhibit certain stationarity properties in time; these include sea-waves, smoke, foliage, whirlwind but also talking faces, traffic etc. We present a novel characterization dynamic poses the problems modelling, learning, recognizing and synthesizing on firm analytical footing. borrow tools from system identification to capture "essence" textures; we do so by learning (i.e. identifying) models optimal sense maximum likelihood or...

Abstract Fractal behavior is scale-invariant and widely characterized by fractal dimension. However, the cor-respondence between them that uniquely determines a dimension while can be related to many possible behaviors. Therefore, independent of generator its geometries, spatial pattern, statistical properties in addition scale. To mathematically describe behavior, we propose novel concept topography defined two parameters, scaling lacunarity ( P ) coverage F ). The as scale ratio successive...

Abstract Pores among particles provide the main space for storage and migration of deep underground fluids (such as oil, gas, groundwater, unconventional natural gas). The pores form a pore structure with complex morphology which is mainly dominated by shape distribution particles. Therefore, reconstruction or granular porous media evaluation particle roundness have become an important foundation study fluid flow through rock mass. This research proposes novel approach multi‐scale model...

The use of the human hand as a natural interface device serves motivating force for research in modeling, analysis and capture motion an articulated hand. Model-based can be formulated large nonlinear programming problem, but this approach is plagued by local minima. An alternative way to analysis-by-synthesis searching huge space, results are rough computation expensive. In paper, decoupled, new two-step iterative model-based algorithm proposed motion, proof convergence also given. our...

Occlusion is a difficult problem for appearance-based target tracking, especially when we need to track multiple targets simultaneously and maintain the identities during tracking. To cope with occlusion explicitly, this paper proposes dynamic Bayesian network, which accommodates an extra hidden process stipulates conditions on image observation likelihood calculated. The statistical inference of such can reveal relations among different targets, makes tracker more robust against partial...

This study proposes a new technique for retrieving temperature and humidity profiles based on Artificial Neural Networks (ANNs) using data acquired from the GIIRS (Geosynchronous Interferometric Infrared Sounder) L1 ERA-Interim (European Centre Medium-Range Weather Forecasts Reanalysis). The approach is also compared against another method that uses simulated radiative transfer model to construct retrieval network. Furthermore, two methods of network construction are evaluated in North China...

The use of Asphalt Plug Joints (APJs) in bridge construction has rapidly expanded due to their distinct advantages driving comfort, shock absorption, noise reduction, and maintenance ease. However, prolonged exposure vehicular loads environmental factors renders this expansion joint susceptible interface effects stress concentration, leading damage, particularly at the layers between pavement joint-cross plate contact areas. To address issues such as study proposes a polyurethane-modified...

The use of human hand as a natural interface device serves motivating force for research in visual analysis highly articulated movement. Since motion covers huge domain, the scope this paper is limited to developments 3D model-based approaches. Numerous models that have been used analyze are studied. Various approaches discussed. Some realistic synthesis methods also included paper. We conclude with some thoughts about future directions.

This article proposes an active basis model and a shared pursuit algorithm for learning deformable templates from image patches of various object categories. In our generative model, template is in the form basis, which consists small number Gabor wavelet elements at different locations orientations. These are allowed to slightly perturb their orientations before they linearly combined generate each individual training or testing example. The can be learned by algorithm. selects sequentially...

This paper introduces a new sparsity prior to the estimation of dense flow fields. Based on this prior, complex field with motion discontinuities can be accurately estimated by finding sparsest representation in certain domains. In addition, stronger additional constraint gradients is incorporated into model cope measurement noises. Robust techniques are also employed identify outliers and refine results. reliably estimate entire from small portion measurements when other corrupted noise....