Training - Tuesday, June 4
Trainings are add-ons and not included in the Symposium Pass registration. Each training session is $25 for USGIF Members and $35 for USGIF Non-Members.
All training sessions are located on the Third Level in Rooms 301A-302C.
Morning Training Sessions
Deep Learning Demystified: AI for GEOINT Practitioners – Booz Allen Hamilton
This training session removes the mental barrier that many people encounter when thinking about AI. It ensures that attendees understand the core concepts of deep learning, enabling them to sort out the realities from the hype. Attendees will better understand how automation through AI will allow them to free up their time for tasks that require human-level understanding and context awareness.
Conflation: Merging Overlapping Map Data – Harris Corporation
How do you combine multiple representations of a geospatial object into a single feature that is geometrically superior with richer attribution? This is the challenge of the conflation process for vector geospatial data. This course will introduce the issues and potential solutions involved with fusing datasets. Participants will learn about procedures for matching and transforming data and will be given overviews of toolsets that exist for automated conflation processing. This course will employ scenarios involving commonly used vector data and attendees will learn methods of choosing the appropriate conflation and validation processes for different types of features.
CANCELED: Introduction to Deep Learning for Computer Vision and GEOINT Applications – MIT Lincoln Laboratory (Room 301A)
Attendees will be shown the essential mathematical underpinnings of supervised deep learning, along with lessons that help build an intuition for the data-driven needs and limitations of learned models. They will also be introduced to the popular open-source technologies most widely-used among deep learning practitioners. Finally, attendees will be provided an overview of the current state of deep learning as it pertains to GEOINT applications and with guidance for GEOINT subject matter experts who are interested in assessing deep learning approaches for solving relevant problems. This session requires a computer.
NGA’s Maritime Support – The National Geospatial-Intelligence Agency (NGA)
Join NGA Maritime leadership to learn about the myriad sources, products, and services that comprise maritime GEOINT. This training will cover NGA’s current Safety of Navigation offerings; their move to cloud-based, data-centric operations; and the transition from MILSPEC products to the global standard – which will increase interoperability and brokering opportunities with our international partners. The session will also cover NGA’s leadership role in developing the next generation of international maritime data standards – for a future of secure machine-to-machine data transfers, integrated data layers, and unmanned vessel navigation.
In-Browser 3D Exploitation of Multi-Platform, Multi-Resolution Imagery – PixElement
This course will provide knowledge and hands-on training on how to rapidly stream, exploit, and disseminate 3D geospatial products directly in the browser. Participants in this workshop will gain understanding of how massive 3D datasets are created from multi-resolution 2D imagery, processed in the cloud, and served on-demand to the client with minimal bandwidth. Specific examples and integrations based on Amazon Web Services and Cesium will be provided.
Unlocking the Power of Persistence with Satellite SAR for Geospatial Intelligence & Discovery – Ursa Space Systems
Attendees will learn the general data flow required for prototyping SAR-based analytics using open-source software and cloud computing. We will discuss extending machine learning, deep learning and transfer learning to SAR data, SAR for interferometric applications, 3D SAR modeling and machine vision, 3D change detection, and data fusion with other types of data, including multispectral imagery, economic indices, high-velocity AIS data, and showcase analysis algorithms and approaches. This course will teach new machine-driven analytics to extract useful intelligence information from multi-modal/multi-temporal SAR data for all-source analysts and to visualize SAR-derived insights in intuitive ways.