CVLGOct 7, 2018

DeepGeo: Photo Localization with Deep Neural Network

arXiv:1810.03077v113 citationsHas Code
Originality Synthesis-oriented
AI Analysis

This addresses the challenge of photo localization for applications in computer vision and learning, though it is incremental as it builds on existing deep learning methods with a new dataset.

The paper tackles the problem of predicting the geographical location of an image by framing it as a classification task across 50 U.S. states, achieving 71.87% accuracy with top-5 guesses and outperforming humans in GeoGuessr.

In this paper we address the task of determining the geographical location of an image, a pertinent problem in learning and computer vision. This research was inspired from playing GeoGuessr, a game that tests a humans' ability to localize themselves using just images of their surroundings. In particular, we wish to investigate how geographical, ecological and man-made features generalize for random location prediction. This is framed as a classification problem: given images sampled from the USA, the most-probable state among 50 is predicted. Previous work uses models extensively trained on large, unfiltered online datasets that are primed towards specific locations. To this end, we create (and open-source) the 50States10K dataset - with 0.5 million Google Street View images of the country. A deep neural network based on the ResNet architecture is trained, and four different strategies of incorporating low-level cardinality information are presented. This model achieves an accuracy 20 times better than chance on a test dataset, which rises to 71.87% when taking the best of top-5 guesses. The network also beats human subjects in 4 out of 5 rounds of GeoGuessr.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes