Which country is this picture from? New data and methods for DNN-based country recognition

TL;DR

This paper introduces a new large dataset and a classification-based model for recognizing the country in which a photo was taken, outperforming previous geo-coordinate estimation methods.

Contribution

The paper presents the VIPPGeo dataset and a novel DNN model for country recognition, demonstrating improved accuracy over existing approaches.

Findings

The model outperforms state-of-the-art geo-coordinate estimation methods.

Country recognition yields better results than coordinate-based inference.

The VIPPGeo dataset contains 3.8 million geo-tagged images for training.

Abstract

Recognizing the country where a picture has been taken has many potential applications, such as identification of fake news and prevention of disinformation campaigns. Previous works focused on the estimation of the geo-coordinates where a picture has been taken. Yet, recognizing in which country an image was taken could be more critical, from a semantic and forensic point of view, than estimating its spatial coordinates. In the above framework, this paper provides two contributions. First, we introduce the VIPPGeo dataset, containing 3.8 million geo-tagged images. Secondly, we used the dataset to train a model casting the country recognition problem as a classification problem. The experiments show that our model provides better results than the current state of the art. Notably, we found that asking the network to identify the country provides better results than estimating the geo-coordinates and then tracing them back to the country where the picture was taken.