103 Hits in 3.0 sec

PlaNet - Photo Geolocation with Convolutional Neural Networks [chapter]

Tobias Weyand, Ilya Kostrikov, James Philbin
2016 Lecture Notes in Computer Science  
Moreover, we extend our model to photo albums by combining it with a long short-term memory (LSTM) architecture.  ...  In computer vision, the photo geolocation problem is usually approached using image retrieval methods.  ...  We then train a convolutional neural network (CNN) [49] using millions of geotagged photos.  ... 
doi:10.1007/978-3-319-46484-8_3 fatcat:5uwderjw5rhcdox5ymflfy4t3i

Geolocating Images with Crowdsourcing and Diagramming

Rachel Kohler, John Purviance, Kurt Luther
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
In this paper, we explore how crowdsourcing can be used to support expert image geolocation.  ...  Professional geolocation is often a manual, time-consuming process that can involve searching large areas of satellite imagery for potential matches.  ...  More recently, PlaNet [Weyand et al., 2016] takes a photo as input and, using a convolutional neural network trained on 126 million geotagged photos from the web, generates a probability for 26,000 cells  ... 
doi:10.24963/ijcai.2018/741 dblp:conf/ijcai/KohlerPL18 fatcat:niphte3mpbchxdsdabhzprc2dy

CPlaNet: Enhancing Image Geolocalization by Combinatorial Partitioning of Maps [article]

Paul Hongsuck Seo, Tobias Weyand, Jack Sim, Bohyung Han
2018 arXiv   pre-print
Image geolocalization is the task of identifying the location depicted in a photo based only on its visual information.  ...  This task is inherently challenging since many photos have only few, possibly ambiguous cues to their geolocation.  ...  Acknowledgment The part of this work was performed while the first and last authors were with Google, Venice, CA. This research is partly supported by the IITP grant [2017-0-  ... 
arXiv:1808.02130v1 fatcat:pd4x6iaknndoxkkzui3ltqs43m

CPlaNet: Enhancing Image Geolocalization by Combinatorial Partitioning of Maps [chapter]

Paul Hongsuck Seo, Tobias Weyand, Jack Sim, Bohyung Han
2018 Lecture Notes in Computer Science  
Image geolocalization is the task of identifying the location depicted in a photo based only on its visual information.  ...  This task is inherently challenging since many photos have only few, possibly ambiguous cues to their geolocation.  ...  Acknowledgment The part of this work was performed while the first and last authors were with Google, Venice, CA.  ... 
doi:10.1007/978-3-030-01249-6_33 fatcat:u25oh5ujzfba7jj6p5lcrfvei4

Revisiting IM2GPS in the Deep Learning Era [article]

Nam Vo, Nathan Jacobs, James Hays
2017 arXiv   pre-print
Image geolocalization, inferring the geographic location of an image, is a challenging computer vision problem with many potential applications.  ...  Training with classification loss outperforms several deep feature learning methods (e.g. Siamese networks with contrastive of triplet loss) more typical for retrieval applications.  ...  PlaNet proposes a deep convolutional neural network to estimate a probability distribution over regions from raw pixel values.  ... 
arXiv:1705.04838v1 fatcat:oqbemxdx7bdmrf7eugsoqcjf5m

Real-time Visual Landmark Recognition in Multi-view Image Collections

Kwisha Hitesh, Sonal Pravinbhai
2019 International Journal of Computer Applications  
Research and advancement in the Convolution Neural Network have been capable of solving many computer vision problems with higher accuracy than humans at some time.  ...  This paper, presents CNN along with its various layers for easy understanding. CNN algorithm has been used here for the landmark recognition problem.  ...  The equation of loss function [10] used is as follows: Convolution Neural Network is a classification technique. It is a type of feed-forward Artificial Neural Network.  ... 
doi:10.5120/ijca2019918922 fatcat:c7xndyejk5fnhevlmukidfcyqu

Regional High-Resolution Benthic Habitat Data from Planet Dove Imagery for Conservation Decision-Making and Marine Planning

Steven R. Schill, Valerie Pietsch McNulty, F. Joseph Pollock, Fritjof Lüthje, Jiwei Li, David E. Knapp, Joe D. Kington, Trevor McDonald, George T. Raber, Ximena Escovar-Fadul, Gregory P. Asner
2021 Remote Sensing  
Results from a combined major class accuracy assessment yielded an overall accuracy of 80% with a standard error of less than 1% yielding a confidence interval of 78%–82%.  ...  Techniques for mapping benthic habitats will also continue to improve with more powerful machine learning classifiers, such as Support Vector Machines and Convolution Neural Networks (CNN) [79, 80] .  ...  A description for each of these classes with corresponding field photo examples can be found in the Supplemental Materials.  ... 
doi:10.3390/rs13214215 fatcat:ajqxdlfqavcjrgst7ucosmq2bu

Interpretable Semantic Photo Geolocation [article]

Jonas Theiner, Eric Mueller-Budack, Ralph Ewerth
2021 arXiv   pre-print
Due to the success of convolutional neural networks (CNNs), current approaches achieve super-human performance. However, previous work has exclusively focused on optimizing geolocalization accuracy.  ...  Planet-scale photo geolocalization is the complex task of estimating the location depicted in an image solely based on its visual content.  ...  ), and train Convolutional Neural Networks (CNNs) with a huge amount of labeled data in an end-to-end fashion.  ... 
arXiv:2104.14995v2 fatcat:j3daqxkg2jbobgmvf6ftuhjimu

Leveraging Selective Prediction for Reliable Image Geolocation [article]

Apostolos Panagiotopoulos, Giorgos Kordopatis-Zilos, Symeon Papadopoulos
2021 arXiv   pre-print
State-of-the-art geolocation methods surpass human performance on the task of geolocation estimation from images.  ...  By abstaining from predicting non-localizable images, we improve geolocation accuracy from 27.8% to 70.5% at the city-scale, and thus make current geolocation models reliable for real-world applications  ...  [5] took advantage of the deep learning advances and trained a Convolutional Neural Network (CNN) to extract features from images.  ... 
arXiv:2111.11952v1 fatcat:q2fvf3hhf5d27fxykyjkagbx7a

Multi-modal Geolocation Estimation Using Deep Neural Networks [article]

Jesse M. Johns, Jeremiah Rounds, Michael J. Henry
2017 arXiv   pre-print
limitations when training these models, and demonstrating how incorporating additional information can be used to improve the overall performance of a geolocation inference model.  ...  This work contributes to the state of research in image geolocation inferencing by introducing a novel global meshing strategy, outlining a variety of training procedures to overcome the considerable data  ...  MODEL The Inception v4 convolutional neural network architecture proposed in (Szegedy et al., 2016) is deployed to develop the mesh-based classification geolocation model (Model M1) presented in this  ... 
arXiv:1712.09458v1 fatcat:ng5hhx7k2nfgrjxxq4z246v6bu

A Multi-Stage Multi-Task Neural Network for Aerial Scene Interpretation and Geolocalization [article]

Alina Marcu, Dragos Costea, Emil Slusanschi, Marius Leordeanu
2018 arXiv   pre-print
We propose a novel multi-task multi-stage neural network that is able to handle the two problems at the same time, in a single forward pass.  ...  Semantic segmentation and vision-based geolocalization in aerial images are challenging tasks in computer vision.  ...  Planet-wide localization of a photo was also investigated [42] , the output being a probability function over the surface of the Earth -with several partitioning schemes being proposed.  ... 
arXiv:1804.01322v1 fatcat:3byzftzvnrc2hegjrujj6zjhau

Where in the World is this Image? Transformer-based Geo-localization in the Wild [article]

Shraman Pramanick, Ewa M. Nowara, Joshua Gleason, Carlos D. Castillo, Rama Chellappa
2022 arXiv   pre-print
In this work, we focus on developing an efficient solution to planet-scale single-image geo-localization.  ...  To this end, we propose TransLocator, a unified dual-branch transformer network that attends to tiny details over the entire image and produces robust feature representation under extreme appearance variations  ...  Recently, convolutional neural networks (CNNs) trained with large datasets have significantly improved the performance of geo-localization methods and enabled extending the task to the scale of the entire  ... 
arXiv:2204.13861v2 fatcat:uphqixmbxjg4baxwja6jdoqbee

Extensive Exposure Mapping in Urban Areas through Deep Analysis of Street-Level Pictures for Floor Count Determination

Gianni Iannelli, Fabio Dell'Acqua
2017 Urban Science  
Acknowledgments: We would like to thank Paolo Gamba at the University of Pavia, Italy, for his support, especially in understanding potentials and limits of Convolutional Neural Networks, and in providing  ...  Figure 2 . 2 Convolutional Neural Network named "VGG-16", partly applied in detecting the number of floors.  ...  These latter requirements match pretty well with the characteristics of methods based on Convolutional Neural Networks (CNN), which is a Deep Learning (DL) technique.  ... 
doi:10.3390/urbansci1020016 fatcat:2tfwyfqbvjfn3ho6tr66namvzq

Visual and Object Geo-localization: A Comprehensive Survey [article]

Daniel Wilson, Xiaohan Zhang, Waqas Sultani, Safwan Wshah
2021 arXiv   pre-print
Philbin, “Planet - 2002. photo geolocation with convolutional neural networks,” [26] H. Zhang, A. Berg, M. Maire, and J.  ...  Fig. 10: The neural network architecture proposed by [34].  ... 
arXiv:2112.15202v1 fatcat:ipwas72ro5ho5fjiakm6de7ji4

Sunrise or Sunset: Selective Comparison Learning for Subtle Attribute Recognition [article]

Hong-Yu Zhou, Bin-Bin Gao, Jianxin Wu
2017 arXiv   pre-print
Experiments show that our approach surpasses baseline methods by a large margin and achieves better results even compared with humans.  ...  The difficulty of image recognition has gradually increased from general category recognition to fine-grained recognition and to the recognition of some subtle attributes such as temperature and geolocation  ...  [27] proposed PlaNet, a deep model to use and integrate multiple visual cues to solve the photo geolocation problem.  ... 
arXiv:1707.06335v1 fatcat:373uavcewbaynnm6ikl2vhvdci
« Previous Showing results 1 — 15 out of 103 results