Estimating the Natural Illumination Conditions from a Single Outdoor Image

Given a single outdoor image, we present a method for estimating the likely illumination conditions of the scene. In particular, we compute the probability distribution over the sun position and visibility. The method relies on a combination of weak cues that can be extracted from different portions of the image: the sky, the vertical surfaces, the ground, and the convex objects in the image. While no single cue can reliably estimate illumination by itself, each one can reinforce the others to yield a more robust estimate. This is combined with a data-driven prior computed over a dataset of 6 million photos. We present quantitative results on a webcam dataset with annotated sun positions, as well as quantitative and qualitative results on consumer-grade photographs downloaded from Internet. Based on the estimated illumination, we show how to realistically insert synthetic 3-D objects into the scene, and how to transfer appearance across images while keeping the illumination consistent.     

A flexible edge matching technique for object detection in dynamic environment

Considering the robustness, stability and reduced volume of data, researchers have focused on using edge information in various video processing applications including moving object detection, tracking and target recognition. Though the edge information is more robust compared to intensity, it also exhibits variations in different frames due to illumination change and noise. In addition to this, the amount of variation varies from edge to edge. Thus, without making use of this variability information, it is difficult to obtain an optimal performance during edge matching. However, traditional edge pixel-based methods do not keep structural information of edges and thus they are not suitable to extract and hold this variability information. To achieve this, we represent edges as segments that make use of the structural and relational information of edges to allow extraction of this variability information. During edge matching, existing algorithms do not handle the size, positional and rotational variations to deal with edges of arbitrary shapes. In this paper, we propose a knowledge-based flexible edge matching algorithm where knowledge is obtained from the statistics on the environmental dynamics, and flexibility is to deal with the arbitrary shape and the geometric variations of edges by making use of this knowledge. In this paper, we detailed the effectiveness of the proposed matching algorithm in moving object detection and also indicated its suitability in other applications like target detection and tracking.     

Generalized Adjusted Rand Indices for cluster ensembles

In this paper, Adjusted Rand Index (ARI) is generalized to two new measures based on matrix comparison: (i) Adjusted Rand Index between a similarity matrix and a cluster partition (ARImp), to evaluate the consistency of a set of clustering solutions with their corresponding consensus matrix in a cluster ensemble, and (ii) Adjusted Rand Index between similarity matrices (ARImm), to evaluate the consistency between two similarity matrices. Desirable properties of ARI are preserved in the two new measures, and new properties are discussed. These properties include: (i) detection of uncorrelatedness; (ii) computation of ARImp/ARImm in a distributed environment; and (iii) characterization of the degree of uncertainty of a consensus matrix. All of these properties are investigated from both the perspectives of theoretical analysis and experimental validation. We have also performed a number of experiments to show the usefulness and effectiveness of the two proposed measures in practical applications.     

Node trust evaluation in mobile ad hoc networks based on multi-dimensional fuzzy and Markov SCGM(1, 1) model

Due to the nature of distribution and self-organization, mobile ad hoc networks rely on cooperation between nodes to transfer information. One of the key factors to ensure high communication quality is an efficient assessment scheme for risks and trust of choosing next potential cooperative nodes. Trust model, an abstract psychological cognitive process, is one of the most complex concepts in social relationships, involving factors such as assumptions, expectations and behaviors. All of the above make it difficult to quantify and forecast trust accurately. In this paper, based on the theories of fuzzy recognition with feedback, SCGM(1, 1) model and Markov chain, we present a pattern of prediction making. The analysis and experimental computation show that this scheme is efficient in trust prediction for ad hoc networks.     

Decision tree classifiers sensitive to heterogeneous costs

There are nine major types of cost involved in cost-sensitive learning that are with heterogeneous units in general, referred to heterogeneous costs. Extant cost-sensitive learning (CSL) algorithms are based on the assumption that all involved costs can be transformed into a unified unit, called as homogeneous assumption of costs. While it is a challenge to construct many suitable transformation functions for the costs with diverse units, this paper designs a heterogeneous-cost sensitive learning (HCSL) algorithm to make split attribute selection more effective. This paper first proposes an efficient method of reducing the heterogeneity caused by both cost mechanisms and attribute information. And then, all heterogeneous costs with attribution information together are incorporated into the process of split attribute selection, called as HCAI-based split attribute selection. Third, the over-fitting is tackled by designing a simple and effective smoothing strategy, so as to build cost-sensitive decision tree classifiers with the HCSL algorithm. Experiments are conducted to evaluate the proposed HCSL algorithm on six UCI datasets. Experimental results show that the proposed approach outperforms existing methods for handling the heterogeneity caused by cost mechanisms and attribute information.     

Efficient screen-space approach to high-quality multiscale ambient occlusion

We present a new screen-space ambient occlusion (SSAO) algorithm that improves on the state-of-the-art SSAO methods in both performance and quality. Our method computes ambient occlusion (AO) values at multiple image resolutions and combines them to obtain the final, high-resolution AO value for each image pixel. It produces high-quality AO that includes both high-frequency shadows due to nearby, occluding geometry and low-frequency shadows due to distant geometry. Our approach only needs to use very small sampling kernels at every resolution, thereby achieving high performance without resorting to random sampling. As a consequence, our results do not suffer from noise and excessive blur, which are common of other SSAO methods. Therefore, our method also avoids the expensive, final blur pass commonly used in other SSAO methods. The use of multiple resolutions also helps reduce errors that are caused by SSAO’s inherent lack of visibility checking. Temporal incoherence caused by using coarse resolutions is solved with an optional temporal filtering pass. Our method produces results that are closer to ray-traced solutions than those of any existing SSAO methods, while running at similar or higher frame rates than the fastest ones.     

Exploring US Business Cycles with Bivariate Loops Using Penalized Spline Regression

The phrase business cycle is usually used for short term fluctuations in macroeconomic time series. In this paper we focus on the estimation of business cycles in a bivariate manner by fitting two series simultaneously. The underlying model is thereby nonparametric in that no functional form is prespecified but smoothness of the functions are assumed. The functions are then estimated using penalized spline estimation. The bivariate approach will allow to compare business cycles, check and compare phase lengths and visualize this in forms of loops in a bivariate way. Moreover, the focus is on separation of long and short phase fluctuation, where only the latter is the classical business cycle while the first is better known as Friedman or Goodwin cycle, respectively. Again, we use nonparametric models and fit the functional shape with penalized splines. For the separation of long and short phase components we employ an Akaike criterion.     

Globally stable adaptive robust tracking control using RBF neural networks as feedforward compensators

In previous adaptive neural network control schemes, neural networks are usually used as feedback compensators. So, only semi-globally uniformly ultimate boundedness of closed-loop systems can be guaranteed, and no methods are given to determine the neural network approximation domain. However, in this paper, it is showed that if neural networks are used as feedforward compensators instead of feedback ones, then we can ensure the globally uniformly ultimate boundedness of closed-loop systems and determine the neural network approximation domain via the bound of known reference signals. It should be pointed out that this domain is very important for designing the neural network structure, for example, it directly determines the choice of the centers of radial basis function neural networks. Simulation examples are given to illustrate the effectiveness of the proposed control approaches.