Urban Perception: Can we understand why a street is safe?

Abstract

The importance of urban perception computing isrelatively growing in machine learning, particularly in relatedareas to Urban Planning and Urban Computing. This fieldof study focuses on developing systems to analyze and mapdiscriminant characteristics that might directly impact the city’sperception. In other words, it seeks to identify and extractdiscriminant components to define the behavior of a city’sperception. This work will perform a street-level analysis tounderstand safety perception based on the “visual components”.As our result, we present our experimental evaluation regardingthe influence and impact of those visual components on the safetycriteria and further discuss how to properly choose confidenceon safe or unsafe measures concerning the perceptional scoreson the city street levels analysis.