Phoenixmap: Spatio-Temporal Distribution Analysis With Deep Learning Classifications

Junhan Zhao, Xiang Liu, Ryan Guan, Josephine Zhang, Baijian Yang, Zhenyu Cheryl Qian, Yingjie Victor Chen

Abstract: In this paper, we introduce PhoenixMap, a visual analytics system to help analysts understand spatiotemporal distribution patterns and anomalies of many objects. It focuses on solving two VA challenges. First, visualizing the spatial distribution of many points under different categories especially when these points partially overlap with each other; Second, employing visual means to classify bird calls using deep learning CNN method. This PhoenixMap system can successfully help us understand the change and patterns of birds regions over many years.