论文标题
脉冲星搜索的公民计算：ciclops
Citizen COmputing for Pulsar Searches: CICLOPS
论文作者
论文摘要
当今Pulsar天文学中使用的大多数周期性搜索算法都非常有效，并利用了多个CPU或GPU。瓶颈通常由需要从专家眼中获得明智选择的操作代表。一个典型的情况是数据中存在射频干扰，通常模仿脉冲星的周期性信号，并且需要目视检查数百或数千个脉冲星“候选者”，以满足许多预选标准。 CICLOPS是一个公民科学项目，旨在将搜索脉冲星的搜索转变为有趣的3D视频游戏。我们构建了一个分布式计算平台，使用用户的CPU和GPU运行计算，并在模式识别中使用独特的人类能力来找到最佳的候选脉动。
Most periodicity search algorithms used in pulsar astronomy today are highly efficient and take advantage of multiple CPUs or GPUs. The bottlenecks are usually represented by the operations that require an informed choice from an expert eye. A typical case is the presence of radio-frequency interferences in the data, that often mimic the periodic signals of pulsars, and require visual inspection of hundreds or thousands of pulsar "candidates" satisfying a number of preselected criteria. CICLOPS is a citizen science project designed to transform the search for pulsars into an entertaining 3D video game. We build a distributed computing platform, running calculations with the user's CPUs and GPUs and using the unique human abilities in pattern recognition to find the best candidate pulsations.
