基于核密度估计多样性采集的背景建模算法

1IntroductionBackground subtraction methods are extensively ex-ploited for moving object detection in videos in manyapplications ,such as traffic monitoring[1],human mo-tion capture[2]and video surveillance[3]. The centralidea behind such methods is to segment foreground ini mage sequences taken from a fixed camera by com-paring each newframe to a model of the scene back-ground . Therefore , correctly and efficiently modelingand updating the background model is one of the mostchallenging aspec…     

Rats remember not wisely but too well

Rats were trained in a three-alternative spatial delayed matching-to-sample task in a starburst maze. Samples consisted of rewarded forced choices of one arm, and retention was indicated by rats’ returning to that arm after a 90-sec delay. If a rat made an error on its first choice, it was returned to the start compartment and allowed a second choice. Unlike in previous experiments with this task, all three arms were available during the animals’ second choices. The rats tended to perseverate in their second choices by returning to the arm that they had erroneously visited on their first choice. In Experiment 1, the accuracy of second choices following first-choice errors was below chance during the first block of sessions, when a 90-sec delay intervened between the first choice and the second choice, and at chance during the second block of sessions, when a short (5–6 see) delay intervened between first and second choices. In Experiment 2, long-delay and short-delay sessions were randomly presented to naive subjects. Similar results were obtained. In both experiments, the tendency to repeat the erroneous first choice was greater when long delays separated the two choices than when short delays were used. The results suggest that rats make their first-choice errors because they erroneously encode or remember the location of the sample and that they base their second choices on the same erroneous-memory. The increase in perseveration at long delays implies some kind of rehearsal-like mechanism that slows forgetting of the memory controlling the first choice.     

Legacies in Black and White: The Racial Composition of the Legacy Pool

Selective universities regularly employ policies that favor children of alumni (known as legacies) in undergraduate admissions. Since alumni from selective colleges and universities historically have been disproportionately white, admissions policies that favor legacies have disproportionately benefited white students. For this reason, legacy policies lead to additional costs in terms of reductions in racial diversity. As larger numbers of minority students graduate from colleges and universities and have children, however, the potential pool of legacy applicants will change markedly in racial composition. This analysis begins with a review of the history and objectives of the preference for children of alumni in undergraduate admissions. We then consider the specific case of the University of Virginia and employ demographic techniques to predict the racial composition of the pool of potential legacy applicants to the university. Significant changes in the racial composition of classes that graduated from the University of Virginia from the late 1960s through the 1970s foreshadow similar changes in the characteristics of alumni children maturing through the next two decades.     

Conversion of CO<Subscript>2</Subscript> to useful substances with composite iron,nickel, and copper catalysts

Composite materials of Fe/Al₂O₃, which consist of small particles of iron supported by thermally stable alumina even at 500–700 °C, have been widely used in the water-gas shift reaction for natural gas reforming. Therefore, Fe/Al₂O₃ is one of the promising candidates for re-transformation of exhausted CO₂ into fuels such as alcohols and hydrocarbons. The development of a CO₂ reforming system using the composite materials of Fe/Al₂O₃ through CO₂ reduction to CO, dissociation of water into hydrogen, and methanol synthesis has been investigated. It was found that dry and steam (i.e. wet) reforming of CO₂ produced almost the same amount of CO. At a temperature above 500 °C, maximal and saturated yields of CO and H₂ from CO₂ and water were obtained. However, this CO₂ reforming system requires higher-pressure conditions from several tens to hundreds standard atmospheric pressure in order to achieve high yield and selectivity for methanol production. In this study we developed the modified CO₂ reforming system by the utilization of Ni and/or Cu instead of Fe in order to obtain other types of useful products such as CO, CH₄, and carbon, more efficiently and selectively under atmospheric pressure. When Ni or Cu was used, conversion of CO₂ was reduced to 76%, while 9% of methane was detected in the case of Ni. On the other hand, though the CO₂ conversion reduced half of the Fe, the selectivity of CO from CO₂ increased to 95% in the case of Cu.     

Numerical investigation of cavitating flow behind the cone of a poppet valve in water hydraulic system

Computational Fluid Dynamics (CFD) simulations of cavitating flow through water hydraulic poppet valves were performed using advanced RNGk-epsilon turbulence model. The flow was turbulent, incompressible and unsteady, for Reynolds numbers greater than 43000. The working fluid was water, and the structure of the valve was simplified as a two dimensional axisymmetric geometrical model. Flow field visualization was numerically achieved. The effects of inlet velocity, outlet pressure, opening size as well as poppet angle on cavitation intensity in the poppet valve were numerically investigated. Experimental flow visualization was conducted to capture cavitation images near the orifice in the poppet valve with 30° poppet angle using high speed video camera. The binary cavitating flow field distribution obtained from digital processing of the original cavitation image showed a good agreement with the numerical result. Project supported by the National Natural Science Foundation of China (No. 59835160) and Scientific Research Foundation for Returned Overseas Chinese Scholars, State Education Ministry of China (No. 50175097)     

Filtration of micro-particles within multi-fiber arrays by adhesive DEM-CFD simulation

A 3D multi-time scale discrete element method-computational fluid dynamic (DEM-CFD) coupling approach was applied to investigate the filtration of micron-sized particles by different types of fiber arrays. Both the pressure drop and the filtration efficiency were examined to indicate the filtration performance of the fiber arrays. Fibers that were uniformly arrayed in a parallel or staggered manner were compared. Results showed that the staggered array showed a better performance than the parallel array in terms of both pressure drop and filtration efficiency. Further, we compared the performance of different staggered arrays, i.e. a regular case, one densified in the front layers and another densified in the back layers. The front densified array was found to enter the clogging and cake filtration stage in the shortest time, leading to the highest filtration efficiency, but the highest pressure drop. The back densified array still achieved a much higher filtration efficiency, despite a much lower pressure drop comparable to that of the regular array. The results suggest that the two kinds of densified arrays may be suited for different purposes, e.g. baghouse filters or breathing masks.     

A novel multi-objective optimization method for the pressurized reservoir in hydraulic robotics

The pressurized reservoir is a closed hydraulic tank which plays a significant role in enhancing the capabilities of hydraulic driven robotics. The spring pressurized reservoir adopted in this paper requires comprehensive performance, such as weight, size, fluid volume, and pressure, which is hard to balance. A novel interactive multi-objective optimization approach, the feasible space tightening method, is proposed, which is efficient in solving complicated engineering design problems where multiple objectives are determined by multiple design variables. This method provides sufficient information to the designer by visualizing the performance trends within the feasible space as well as its relationship with the design variables. A step towards the final solution could be made by raising the threshold on performance indicators interactively, so that the feasible space is reduced and the remaining solutions are more preferred by the designer. With the help of this new method, the preferred solution of a spring pressurized reservoir is found. Practicability and efficiency are demonstrated in the optimal design process, where the solution is determined within four rounds of interaction between the designer and the optimization program. Tests on the designed prototype show good results.