Deep Learning models for passability detection of flooded roads (original) (raw)

In this paper we study and compare several approaches to detect floods and evidence for passability of roads by conventional means in Twitter. We focus on tweets containing both visual information (a picture shared by the user) and metadata, a combination of text and related extra information intrinsic to the Twitter API. This work has been done in the context of the MediaEval 2018 Multimedia Satellite Task. 2 RELATED WORKS Recent literature approaches leverage on satellite [8, 12, 15] or ground acquisitions [5] to identify flood events. Other works focus more on urban elements detection such as roads [6, 9]. To the best of our knowledge there are no existing works to determine road passability evidence during flood events. 3 DATA The dataset used in this work was distributed by MediaEval 2018 Multimedia Satellite Task [1, 4]. It consists of 5820 Twitter images with its related metadata, from which ∼36% of the images present flooded regions with evidence of roads. Only the images belonging to the earlier class are considered for the second task evaluation: among them, the ∼45% present passable roads. Furthermore, for Copyright held by the owner/author(s).