基于部件模型的无人机系统小样本车辆目标识别方法

对地目标检测与识别是无人机系统典型任务之一,但受限于任务特殊性,往往难以获取足够的目标样本数据以实现高可靠的目标识别.为此,结合人的认知特性,提出一种基于部件模型的小样本车辆目标识别方法,可有效提高无人机感知能力.采用视觉显著性检测与物体性检测相结合的检测方法,提取目标可能区域;采用基于图论的GrabCut方法与最大类...     

Systems failure: the US way of irregular warfare

ABSTRACT

Since 9/11, the United States has achieved notable gains against al Qaeda, and also Islamic State (IS), all while avoiding another mass-casualty attack at home. Yet, institutionally, culturally, and in its capabilities, the US government remains seriously ill-equipped for the task of countering irregular threats. Partly as a result, Islamist extremism shows no sign of being defeated, having instead metastasized since 9/11 and spread. Why, given the importance accorded to counterterrorism, has the US approach remained inadequate? What is impeding more fundamental reforms? The article evaluates the United States’ way of irregular warfare: its troubled engagement with counterinsurgency and its problematic search for lower cost and lower risk ways of combating terrorism. It suggests needed reforms but acknowledges also the unlikelihood of change.