YOLSO: You Only Look Small Object

Small object detection is challenging and far from satisfactory. Most general object detectors suffer from two critical issues with small objects: (1) Feature extractor based on classification network cannot express the characteristics of small objects reasonably due to insufficient appearance information of targets and a large amount of background interference around them. (2) The detector requires a much higher location accuracy for small objects than for general objects. This paper proposes an effective and efficient small object detector YOLSO to address the above problems. For feature representation, we analyze the drawbacks in previous backbones and present a Half-Space Shortcut(HSSC) module to build a background-aware backbone. Furthermore, a coarse-to-fine Feature Pyramid Enhancement(FPE) module is introduced for layer-wise aggregation at a granular level to enhance the semantic discriminability. For loss function, we propose an exponential L1 loss to promote the convergence of regression, and a focal IOU loss to focus on prime samples with high classification confidence and high IOU. Both of them significantly improves the location accuracy of small objects. The proposed YOLSO sets state-of-the-art results on two typical small object datasets, MOCOD and VeDAI, at a speed of over 200 FPS. In the meantime, it also outperforms the baseline YOLOv3 by a wide margin on the common COCO dataset.     

Prediction of risk factors of cyberbullying-related words in Korea: Application of data mining using social big data

The study examined a decision tree analysis using social big data to conduct the prediction model on types of risk factors related to cyberbullying in Korea. The study conducted an analysis of 103,212 buzzes that had noted causes of cyberbullying and data were collected from 227 online channels, such as news websites, blogs, online groups, social network services, and online bulletin boards. Using opinion-mining method and decision tree analysis, the types of cyberbullying were sorted using SPSS 25.0. The results indicated that the total rate of types of cyberbullying in Korea was 44%, which consisted of 32.3% victims, 6.4% perpetrators, and 5.3% bystanders. According to the results, the impulse factor was also the greatest influence on the prediction of the risk factors and the propensity for dominance factor was the second greatest factor predicting the types of risk factors. In particular, the impulse factor had the most significant effect on bystanders, and the propensity for dominance factor was also significant in influencing online perpetrators. It is necessary to develop a program to diminish the impulses that were initiated by bystanders as well as victims and perpetrators because many of those bystanders have tended to aggravate impulsive cyberbullying behaviors.     

Twining Poly(polyoxometalate) Chains into Nanoropes

Organic polymers and inorganic clusters belong to two different disciplines and have completely different properties and structures. When a cluster is attached to the backbone of a polymer as a pendant, the resultant hybrid polymers (polyclusters) exhibit unique behaviours totally different from those of conventional polymers owing to the nanoscale size of the cluster and its particular interactions. Herein, the aggregation of a poly(polyoxometalate)—a polynorbornene backbone with inorganic polyoxometalate cluster pendants—upon addition of a non-solvent to its dilute solution is reported. A three-dimensional network of tangled and snake-like nanothreads was observed. Direct visualisation of individual nanoscale clusters enabled identification of single chains within the nanothreads. These observations suggest that during the process of aggregation, the hybrid polymer forms curved or extended chains as a consequence of an armouring effect in which the collapsed cluster pendants wrap around the backbone. The collapse occurs because they become less soluble in the solvent/non-solvent mixture. The extended chains then become entwined and form nanoropes consisting of multiple chains wound around each other. This study provides a deeper understanding of the nature of polyclusters and should also prove useful for their future development and application.     

A Novel Approach to Analyzing the Occupancy of the Geo/Geo/∞ Queueing System

Discrete time queueing systems have been successfully used to model packet queues in computer and digital communications systems. In this article we introduce a novel approach to analyzing the Geo^X/Geo^Y/∞ queueing system, which is suitable for modelling high speed computer and communications systems. The approach is based on characterizing the system by a functional equation which is then solved recursively to give identical results to those obtained using other approaches.     

Attribute Allocation and Retrieval Scheme for Large-Scale Sensor Networks

Wireless sensor network is an emerging technology that enables remote monitoring of large geographical regions. In this paper, we address the problem of distributing attributes over such a large-scale sensor network so that the cost of data retrieval is minimized. The proposed scheme is a data-centric storage scheme where the attributes are distributed over the network depending on the correlations between them. The problem addressed here is similar to the Allocation Problem of distributed databases. In this paper, we have defined the Allocation Problem in the context of sensor networks and have proposed a scheme for finding a good distribution of attributes to the sensor network. We also propose an architecture for query processing given such a distribution of attributes. We analytically determine the conditions under which the proposed architecture is beneficial and present simulation results to demonstrate the same. To the best of our knowledge, this is the first attempt to determine an allocation of attributes over a sensor network based on the correlations between attributes.     

构建高扩展性的县级广电宽带数据网络

提出在县市级广电宽带数据网络的建设过程中,要着眼业务发展,突出高扩展性,构建基于标准、扩展灵活的主干网,实现网络建设的可持续性发展,并结合实际,从核心设备、端口、中继带宽、网络业务4个方面对扩展能力进行了阐述.     

USERS' TWO-DIMENSIONAL INTEREST AND ITS STORAGE METHOD

The users' interest can be mined from the web cache and can be used widely. The interest can be specialized by the two-tuple (term, weight) in the simple interest model, in which the association relations are not mined, and then the interest cannot be associated in expressing the users' interest. Based on analyzing the WWW cache model, this letter brings forward a two-dimensional interest model and gives the interrelated methods on how to store the two-dimensional interest model effectively.     

USERS’ TWO-DIMENSIONAL INTEREST AND ITS STORAGE METHOD

The users‘ interest can be mined from the web cache and can be used widely.The interest can be specialized by the two-tuple(term,weight) in the simple interest model,in which the association relations are not mined.and then interest cannot be associated in expressing the users‘interest.Based on analyzing the WWW cache model,this letter brings forward a twodimensional interest model and gives the interrelated methods on how to store the two-dimensional interest model effectively.     

并行哈夫曼编码器的硬件设计与实现

文章设计了一种并行编码的哈夫曼硬件编码器，它采用了流水线和并行编码方法，使得在一个时钟周期内可以编码一个字节的数据，在编码时显著降低了工作频率。文章给出了关键部分的实现方案并分析了实验结果。