Award Data

In Phase I, our team of Kitware and UC-Berkeley developed Scale-MAE by adding ground sample distance (GSD) to positional encodings, and produced a multiscale representation that achieves state-of-the art results across image classification, semantic segmentation, and object detection tasks. In Phase II, we will create a remote sensing pretraining toolkit to enable fast and easy experimentation wit ...

Many municipalities in the United States have begun to use impermeability measurements to incentivize storm water retention as part of the architectural and landscaping design process. These measurements are performed through analysis of high-resolution aerial photography and updated on an annual or semi-annual basis. The practice of using high resolution aerial or even satellite photography is pr ...

The NGA program “GLIMPSE” leverages context and topography to geolocate imagery for further analysis. For this proposal, Clarifai intends to develop and deliver a deep-learning pipeline to reduce the geographical search space for an image that will expedite analysis and reduce computational cost. The objective of this proposal is to provide a system to (a) efficiently identify relevant imagery ...

Object detection datasets for overhead imagery are typically generated using bootstrapping methods to reduce annotator effort and cost. These methods iteratively train a detector from a limited set of user-provided and model-predicted labels. Such approaches bias detectors toward the initial object set, limiting their capacity to handle object variations or discover novel objects classes. MONET ov ...

NGA uses deep networks for many tasks including image registration, land cover segmentation, and object detection. Current deep learning approaches develop specialist networks for each task and type of data. Not only is this inefficient, because networks can’t be reused across tasks, this approach ignores correlations between tasks and data sources that can improve performance. In response, we w ...

Nanoplastics (plastic debris

Seafood consumption in the United States has reached over 50% per capita over the last 25 years with over 75% of total U.S. seafood consumption coming from import. Many of the U.S. seafood import sources are located in tropical areas where bacteria and toxin hazards thrive. This concern is exacerbated by the increasing trade deficit. In 2001, the United States imported $6.8 billion more than expor ...

The primary goal of this research is to advance flood extent delineation modeling by enhancing the ability of the modeling community to access and utilize the forecasts and reanalysis products of the National Water Model. One of the most difficult challenges facing stakeholders in the flood inundation domain is the lack of easily accessible tools to query and process the large and complex datasets ...

One of the most pressing problems at the forefront of machine learning and computer vision research is image recognition when high quality labeled data is difficult to acquire, typically due to the prohibitive cost of data acquisition and annotation. Clarifai has teamed with L3Harris to combine their unparalleled experience in simulating geospatial sensors with our award-winning computer vision ca ...

For this proposal, Clarifai would utilize internal expertise in computer vision and deep learning to pursue a CNN to camera orientation estimation. Specifically, we would adapt existing Clarifai models to extract image features and use these features as input to a camera orientation regressor. The horizon filter information might be combined with information from the encoder to produce a joint emb ...