Data Sheets

MXR-DS-geohive 06/20 When clients need to confirm the presence of a feature or features, GeoHIVE validation campaigns utilize the skills of the crowd to quickly assess the presence or attributes of a feature within a specified polygon. These campaigns can ask the crowd to confirm or deny the presence of a bridge within a circle or to appropriately label a recreational point of interest as a swimming pool, park, or soccer field. We attain high-quality results by asking several crowd members to vote on each feature. When the crowd has completed the campaign, CrowdRank® calculates the most likely classification and confidence score of each user and feature before our GeoHIVE analysts perform the final QC checks. GeoHIVE in action: Functional map of the world info@maxar.com maxar.com delivery methods ■ ESRI file geodatabase ■ Shapefile ■ KML ■ CSV ■ GeoJSON ■ Custom analytics report ■ FTP delivery markets ■ Defense and intelligence agencies ■ Humanitarian assistance and disaster response Insurance and reinsurance ■ Defense and homeland security ■ Oil and gas exploration ■ Infrastructure and supply chain monitoring ■ Commodities investors Captions map showing density of features validated by GeoHIVE in support of the functional map of the world challenge. Validate Create When clients are interested in mapping an area, our GeoHIVE Editor campaigns are ideal for accurately digitizing and mapping an AOI in a short time frame. First, we divide the AOI into many rectangular sections, then make sure each section is viewed by one or more crowd members. Users then create point, line, and polygon features, as well as assign attributes such as land use or name. Once an image is digitized, it is sent to another user to validate and approve the bounded/attributed feature(s). This capability is typically reserved for university outreach or internal dataset enrichment.GeoHIVE Editor campaigns are recorded on a proprietary server, allowing the crowd to digitize temporary features and events, making it one of the most powerful crowdsourcing platforms in the world. In 2017, Maxar was part of a revolutionary machine learning dataset project called Functional Map of the World. It is comprised of 1,000,000 labeled satellite image chips within 85 distinct object types. GeoHIVE supported the effort by validating over 450,000 points of interest from open source data and creating a feature bounding box for each image chip. Labeled satellite image chip with a soccer field bounding box drawn by the GeoHIVE crowd in support of the Functional Map of the World challenge

• Cover