For shot boundary detection, our approach combines pairwise similarity analysis and supervised classification. Using primitive low-level image features, we build secondary features based on inter-frame dissimilarity. The secondary features are motivated by prior work on media segmentation in which a kernel function is correlated along the main diagonal of a similarity matrix to construct a fram...

Play is often episodic and mission-centric, with a series of challenges culminating in a final puzzle or enemy that must be overcome. Multiple missions played with the same characters may be related to each other in a plot arc of escalating challenges. The exact tone, structure, pace and end (if any) vary from game to game depending on the needs and preferences of the players, as in [9]. "THE C...

We consider improving the performance of k-Nearest Neighbor classifiers. A regularized kNN is proposed to learn an optimal dissimilarity function to substitute the Euclidean metric. The learning process employs hyperkernels and shares a similar regularization framework as support vector machines (SVM). Its performance is shown to be consistently better than kNN, and is competitive with SVM.

The present study aimed to determine the composition and structure of fringe inland mangroves near mouth Amazon River in Marajó Island. We established five sites and, each site, we a zone an zone. In zone, delimited 400-m2 plots, selected arboreal individuals with minimum height 1 m. were grouped into diameter classes basal area, analyzed phytosociological parameters. composition, richness, den...

A common approach in structural pattern classification is to define a dissimilarity measure on patterns and apply a distance-based nearest-neighbor classifier. In this paper, we introduce an alternative method for classification using kernel functions based on edit distance. The proposed approach is applicable to both string and graph representations of patterns. By means of the kernel function...