• Publications
  • Influence
DEEPFOCAL: A method for direct focal length estimation
This work explores the application of a deep convolutional neural network, trained on natural images obtained from Internet photo collections, to directly estimate the focal length using only raw pixel intensities as input features. Expand
Large-scale geo-facial image analysis
GeoFaces, a large dataset of geotagged face images, is constructed and used to examine the geo-dependence of facial features and attributes, such as ethnicity, gender, or the presence of facial hair, both globally and in selected major urban areas. Expand
A fast method for estimating transient scene attributes
This work proposes the use of deep convolutional neural networks to estimate the transient attributes of a scene from a single image, and shows how this method is more accurate and significantly faster than previous methods, enabling real-world applications. Expand
Learning to Map Nearly Anything
This work proposes a cross-modal distillation strategy to learn to predict the distribution of fine-grained properties from overhead imagery, without requiring any manual annotation of overhead imagery. Expand
Extending Absolute Pose Regression to Multiple Scenes
A novel architecture is proposed, Multi-Scene PoseNet (MSPN), that allows for a single network to be used on an arbitrary number of scenes with only a small scene-specific component, and achieves competitive performance for two bench-mark 6DOF datasets, Microsoft 7Scenes and Cambridge Landmarks. Expand
Learning Geo-Temporal Image Features
We propose to implicitly learn to extract geo-temporal image features, which are mid-level features related to when and where an image was captured, by explicitly optimizing for a set of location andExpand
GeoFaceExplorer: exploring the geo-dependence of facial attributes
The images uploaded to social networking websites are a rich source of information about the appearance of people around the world. We present a system, GeoFaceExplorer, for collecting, processing,Expand
What Goes Where: Predicting Object Distributions from Above
In this work, we propose a cross-view learning approach, in which images captured from a ground-level view are used as weakly supervised annotations for interpreting overhead imagery. The outcome isExpand
Weakly-Supervised Feature Learning via Text and Image Matching
The key idea is to use a contrastive loss to train image and text feature extractors to recognize if a given image-finding pair is a true match, then fine-tuned, in a transfer learning setting, for a supervised classification task. Expand
Implicit Land Use Mapping Using Social Media Imagery
This work argues that the abstract notion of land use can be indirectly characterized by the types and quantities of common objects found in an area and proposes an implicit approach to defining and estimating land use that relies on sparsely distributed social media imagery but retains the benefits of dense coverage provided by satellite imagery. Expand