Abstract In video behavior recognition, making full use of the spatio-temporal information contained in frame is critical point to further improve recognition accuracy. our study, we propose a algorithm based on two-stream heterogeneous network, so that can extract feature frame, and finally take advantage features sequence. view characteristics RGB optical flow images, this paper uses DenseNet...

There can be no doubt that nanotechnology will play a major role in our futuretechnology. Computer science offers more opportunities for quantum andnanotechnology systems. Soft Computing techniques such as swarm intelligence, canenable systems with desirable emergent properties. Optimization is an important anddecisive activity in structural designing. The inexpensive re...

this paper considers the job scheduling problem in virtual manufacturing cells (vmcs) with the goal of minimizing two objectives namely, makespan and total travelling distance. to solve this problem two algorithms are proposed: traditional non-dominated sorting genetic algorithm (nsga-ii) and knowledge-based non-dominated sorting genetic algorithm (kbnsga-ii). the difference between these algor...

background: high-risk unsafe behaviors (hrubs) have been known as the main cause of occupational accidents. considering the financial and societal costs of accidents and the limitations of available resources, there is an urgent need for managing unsafe behaviors at workplaces. the aim of the present study was to find strategies for decreasing the rate of hrubs using an integrated approach of s...

routing in ad-hoc networks, specifically intelligent-based ones, is a highly interested research topic in recent years. most of them are simulation-based study. large percentages have not even mentioned some of the fundamental parameters. this strictly reduces their validity and reliability. on the other hand, there is not a comprehensive framework to classify routing algorithms in wireless sen...