Special Issue Editorial: Computational Intelligence and Applications

Abstract

This special issue includes seven highly selected papers from the Fourth International Workshop on Advanced Computational Intelligence (IWACI 2011), held from October 19 to 21, 2011, in Wuhan, China. The IWACI 2011 was a great success and provided a high-level international forum for scientists and engineers to present the latest research in the filed of computational intelligence and applications. To highlight the success of this conference, we edited this special issue for the Cognitive Computation. We chose seven papers from 412 papers submitted to the IWACI’11, as these papers reflect the high quality of the presentations at the conference while capturing the spirit of our theme, ‘‘Computational Intelligence and Applications,’’ for this special issue. Over the past decades, we have witnessed tremendous efforts and developments from all aspects of computational intelligence research, ranging from theoretical foundations, principles, algorithms, to practical applications in different domains. To reflect a flavor of such recent research activities in the community, we carefully selected these seven papers for this special issue, with a focus on different computational intelligence techniques and real-world applications. The first group of papers is focused on the computational intelligence techniques for image processing. As computer vision and image processing has become a critical issue for many of today’s data-intensive applications, ranging from civilian applications such as video processing to surveillance and security-related applications, we selected three papers on this topic. The paper by Zhao et al. discussed an important method for non-blind image de-blurring from a single image. In order to overcome the limitations of the traditional maximum a posteriori (MAP) estimation for nonblind image de-blurring, the authors proposed a Bayesian minimum mean squared error (MMSE) estimation to perform de-blurring. This method is based on an effective statistical framework to model prior knowledge of natural images with an efficient Gibbs sampling algorithm. Various simulations and comparative studies have been developed to demonstrate the effectiveness of this approach. In the paper by Ding et al., a method based on neural network and selforganizing map (SOM) was proposed to connect a sequence of images acquired by a rotating camera. This is a very interesting problem with broader practical applications. A detailed algorithm with a system-level implementation has been presented in this paper. Image classification was studied in the paper by Bian et al. Specifically, the authors presented a method based on modified k-means support vector machines (SVMs) with hybrid sampling method for image classification. This method can reduce convex hulls of clustering structure as well as the overtraining caused by active sampling approaches. Simulation results based on three challenging remote sensing images with comparative study of random sampling (RS) and margin sampling (MS) demonstrated the effectiveness of this approach. The second group includes two theoretically focused papers. In the paper by Feng et al., an interesting problem of stochastic suppression and stabilization of nonlinear Z. Zeng Department of Control Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China e-mail: zgzeng@gmail.com