Pedestrian trajectory prediction is a basic task in computer vision field. The prosperity of artificial intelligence makes the automatic drive, human-robot interaction and surveillance video attract a great deal of attention. Generally, researchers always place emphasis on pedestrian trajectory. The focuses of pedestrian trajectory prediction task are motion pattern modelling and spatio-temporal interaction modelling in the current study. In our paper, we present a GAN-based framework to model pedestrian motion pattern. A Delaunay triangulation algorithm is applied to map the pedestrian interaction. From the perspective of space, both the position interaction and motion interaction of pedestrians can be considered. For example, the influence of the movement direction and motion potential energy of pedestrians on the surrounding pedestrians can be modelled.

No abstract

No abstract

Feature selection are highly important to improve the classification accuracy of recognition systems for foreign matter in cotton. To address this problem, this paper presents six filter approaches of feature selection for obtaining the good feature combination with high classification accuracy and small size, and make comparisons using support vector machine and k-nearest neighbor classifier. The result shows that filter approach can efficiently find the good feature sets with high classification accuracy and small size, and the selected feature sets can effectively improve the performane of recognition system for foreign matter in cotton. The selected feature combination has smaller size and higher accuracy than original feature combination. It is important for developing the recognition systems for cotton matter using machine vision technology.