In this paper, we make use of a multiple-kernel scheme as a guide to exploit the transform kernel selection strategy of the hybrid video coding. We have found that the kernel selection tendency lies in a feature which can be extracted from a pair of kernels. We propose using the IK(1,2,1) kernel for Iand P-Frames and the IK(5,7,3) kernel for B-Frames for the H.264 Video Coding Standard. This gi...

Many progressive wavelet-based image coders are designed for good performance on natural images. They attempt to achieve the greatest reduction in mean squared error (MSE) with each bit sent, an approach that is most effective when the image is composed chiefly of low-frequency content. Many images, however, include sharp-edged objects, text characters or graphics that are not well handled by s...

To test the hypothesis that correlated neuronal activity serves as the neuronal code for visual feature binding, we applied information theory techniques to multiunit activity recorded from pairs of V1 recording sites in anaesthetised cats while presenting either single or separate bar stimuli. We quantified the roles of firing rates of individual channels and of cross-correlations between reco...

This paper presents a compression scheme for color images, by using Self-Organizing Feature Map algorithm which is a neural network structure. In this application 1-dimensional SOFM is used to map 256-color to 64-, 32and 16-color. After the quantization process, relative coding and entropy coding are performed without any loss in the information.

In this paper we apply three pattern recognition methods (support vector machine, cluster analysis and principal component analysis) to distinguish regulatory regions from coding and non-coding non regulatory DNA sequences. Using a new feature representation (the degree by which motifs are overand under-represented) we demonstrate the remarkable power of this methodology in identifying regulato...

In this report, we discuss a content-based video coding approach, i.e. video primal sketch. Unlike traditional video coding approaches, video primal sketch seeks to capture different mid-level cues and use them to explain the video data. The algorithm first segmented the video data into two different kinds of regions: explicit region and implicit region, and then model each region by a correspo...

We propose a new framework for learning the world dynamics of feature-rich environments in model-based reinforcement learning. The main idea is formalized as a new, factored state-transition representation that supports efficient online-learning of the relevant features. We construct the transition models through predicting how the actions change the world. We introduce an online sparse coding ...

Deriving from the gradient vector of a generative model of local features, Fisher vector coding (FVC) has been identified as an effective coding method for image classification. Most, if not all, FVC implementations employ the Gaussian mixture model (GMM) to characterize the generation process of local features. This choice has shown to be sufficient for traditional low dimensional local featur...

Bottleneck autoencoders have been actively researched as a solution to image compression tasks. However, we observed that bottleneck autoencoders produce subjectively low quality reconstructed images. In this work, we explore the ability of sparse coding to improve reconstructed image quality for the same degree of compression. We observe that sparse image compression produces visually superior...

Feature extraction, coding and pooling, are important components on many contemporary object recognition paradigms. In this paper we explore novel pooling techniques that encode the second-order statistics of local descriptors inside a region. To achieve this effect, we introduce multiplicative second-order analogues of average and maxpooling that together with appropriate non-linearities lead ...