Training_Data

Founded in 2015 as part of IQT Labs, CosmiQ Works launched to focus on the geospatial analytics market and provide technical insights, targeted research, reports, and more. Over the past six years CosmiQ has produced many projects and insights that have helped the intelligence and academic communities better understand how geospatial can help tackle hard problems. Our final Training_Data podcast brings together CosmiQ's current colleagues and alum in a special episode celebrating the team's key accomplishments, milestones, and lessons learned through the years.

Satellite imagery analytics have numerous human development and disaster response applications, particularly when time series methods are involved. The Multi-Temporal Urban Development SpaceNet 7 Challenge focuses on developing novel computer vision methods for non-video time series data, asking participants to identify and track buildings in satellite imagery time series collected over rapidly urbanizing areas. In this episode, CosmiQ’s Ryan Lewis, Adam Van Etten, and Daniel Hogan are joined by Planet’s Jesus Martinez Manzo and AWS Disaster Response’s Grace Kitzmiller to explore this new challenge.Learn more at www.spacenet.ai, and at the DownLinQ (https://medium.com/the-downlinq)

CosmiQ’s Jake Shermeyer and Daniel Hogan are joined by Capella Space’s Jason Brown and IEEE Geoscience and Remote Sensing’s (GRSS) Ronny Hänsch to once again discuss the SpaceNet 6 Dataset and post-challenge experiments. Learn more about data fusion and deep learning approaches that work to blend synthetic aperture radar (SAR) and optical imagery. Additionally, the podcast also explores the value of frequent SAR revisits that can be beneficial for foundational mapping applications. Finally, the group discusses CosmiQ’s new addition to the Solaris python package: a multi-modal pre-processing library and a new API called ‘PipeSegment’ for seamlessly stringing together different operations.SpaceNet is a nonprofit made possible by co-founder and managing partner, CosmiQ Works; co-founder and co-chair, Maxar Technologies; and all the other Partners: Amazon Web Services (AWS), Capella Space, Topcoder, IEEE Geoscience and Remote Sensing (GRSS), the National Geospatial-Intelligence Agency, and Planet.  

SpaceNet is a non-profit dedicated to accelerating open source, applied research in geospatial machine learning. In this episode, CosmiQ’s Ryan Lewis, Jake Shermeyer, and Daniel Hogan discuss the SpaceNet 6 Challenge where participants were asked to automatically extract building footprints with computer vision and AI algorithms using a combination of synthetic aperture radar (SAR) and electro-optical imagery. Hear about the challenge’s winning artificial intelligence models and the tradeoff between inference speed and model performance. SpaceNet is made possible by co-founder and managing partner, CosmiQ Works; co-founder and co-chair, Maxar Technologies; and all the other Partners: Amazon Web Services (AWS), Capella Space, Topcoder, IEEE Geoscience and Remote Sensing (GRSS), the National Geospatial-Intelligence Agency, and Planet.Learn more at www.spacenet.ai, and at the DownLinQ (https://medium.com/the-downlinq)

Despite its application to myriad humanitarian and civil use cases, automated road network extraction from overhead satellite imagery remains quite challenging. However, the SpaceNet 5 challenge made significant progress in this field with top participants being able to extract both road networks and speed/travel time estimates for each roadway. On today’s pod, CosmiQ’s Ryan Lewis and Dr. Adam Van Etten explore the challenge’s unique dataset and geographic diversity over time, the winning models, and the tradeoff between inference speed and model performance. SpaceNet is a non-profit LLC co-founded and managed by In-Q-Tel's CosmiQ Works in collaboration Maxar Technologies, a co-founder, and the other SpaceNet Partners including AWS, Intel AI, Topcoder, Capella Space, IEEE GRSS, The National Geospatial-Intelligence Agency (NGA), and Planet.

How can lessons from geospatial computer vision applications impact bio image analysis? CosmiQ’s Dr. Nick Weir and B.Next’s Dr. Dylan George explore the intersection of these two fields and why artificial intelligence (AI) has struggled to gain traction with both satellite imagery and medicine. Hear about their project that researched similarities and differences between satellite imagery, microscopy, and “normal” photographs, and why researchers developing AI methods for microscopy might want to leverage the work done for geospatial applications.Read more about their project at The DownLinQ and BioQuest:· Viewing the World Through a Straw, Part 1· Viewing the World Through a Straw, Part 2

How does emerging technology like artificial intelligence (AI) fit into the broader geopolitical landscape, particularly one that continues to rapidly evolve? CSIS’s Melissa Dalton and Lindsey Sheppard join the podcast to discuss their recent research project, the Gray Zone Project (https://www.csis.org/grayzone), on current challenges faced by the U.S. and its allies, and how emerging technologies could be leveraged to help resolve or mitigate some of the analytic tasks facing national security organizations.CSIS is an independent, bipartisan think tank in Washington, D.C. focused on producing objective analytic analysis that is targeted primarily toward U.S policy makers, but also an increasingly broader audience to include the tech community, allies and partners, and private sector. CSIS’s International Security Program analyzes how the U.S. can best deter, campaign in, and respond to “gray zone” approaches.Learn more about CosmiQ at www.cosmiqworks.org, and CSIS at www.csis.org.

How much labeled data does a machine learning model need in order to achieve sufficient performance? Most analyses simply use all available data and focus on model architecture, with scant attention given to whether the dataset size is appropriate for the task and architecture’s complexity. But is that the right answer?In this episode, IQT CosmiQ Works’ Ryan Lewis, Daniel Hogan, and Adam Van Etten dive into the importance of deep quantitative analysis and the impact of training data size on model performances. Learn how they used CosmiQ’s Machine Learning Robustness Study to help address dependence on geography and dataset size at the leading edge of geospatial machine learning.Learn more about CosmiQ at www.cosmiqworks.org.

Spanning three years of featuring five unique datasets and challenges, SpaceNet® continues to apply various aspects of machine learning to solve difficult foundational mapping problems. Looking ahead, Payam Banazadeh, Founder & CEO of Capella Space, joins IQT CosmiQ’s Ryan Lewis and Jake Shermeyer to discuss the SpaceNet 6 Challenge which explores an under-explored modality of data: Synthetic Aperture Radar (SAR). The trio provide an overview on the importance of SAR data and its value in the upcoming challenge. SpaceNet is a collaboration between CosmiQ Works, Maxar Technologies, Amazon Web Services (AWS), Intel AI, Capella Space, Topcoder, and IEEE GRSS. Space Club Rule 42: When In Doubt, Use SAR.

Thank you to everyone who made the inaugural season of Training_Data possible! We will be back with new content for Season 2 in January 2020. Have a great holiday season and happy new year.

Recorded LIVE at the TopCoder Open 2019 in Houston, Texas! Topcoder’s Global Director for Data, Analytics, & Artiificial Intelligence, Andy LaMora, and General Manager for Public Sector, Michael Contreras, join IQT CosmiQ’s Ryan Lewis and Adam Van Etten at the Topcoder Open 2019, the annual programming and design tournament that took place on November 13 -16, 2019. Learn how Topcoder has evolved over the past almost 20 years, including their involvement with SpaceNet 5 which challenged participants to automatically extract road networks and routing information from satellite imagery.

John Sipple, Senior Software Engineer for Machine Learning at Google and the Artificial Intelligence (AI) Portfolio Joint Reserve Lead for the Defense Innovation Unit, joins IQT CosmiQ Works’ Ryan Lewis and Nick Weir to share lessons learned from designing and executing machine learning projects. The group dives into Sipple’s past experiences, current research areas, and future AI trends and areas of interest, such as attention networks.

Ryan Ashley and John Speed, from IQT’s cybersecurity lab (Cyber Reboot), join IQT CosmiQ’s Ryan Lewis to discuss outcomes from Hacktoberfest, a month-long celebration of open source software. Hear how Cyber Reboot led Hacktoberfest efforts at In-Q-Tel (IQT) to create awareness about and contribute to IQT’s open source projects and beyond. The group also discusses upcoming cybersecurity projects and what to look out for next!

CARTO’s Chief Strategy Officer, Javier de la Torre, and Director of Spatial Data Science, Matt Forrest, join IQT CosmiQ Works’ Ryan Lewis and Nick Weir to discuss how spatial analytics tools can be designed for a variety of different end users, ranging from data scientists to business analysts. They also consider how historically GIS functions are being incorporated in business intelligence (BI) tools. The group closes on a conversation about the Spatial Data Science Conference which was hosted on October 16th at Columbia University.

Special guest Will Porteous, General Partner and COO at RRE Ventures, joins host Ryan Lewis, IQT CosmiQ, and Tom Gillespie, a Managing Partner on IQT’s investment team, to discuss investment strategies, technology trends and predictions, and what’s next for the commercial space and aerospace industry. Commercial space and, more broadly, Frontier Technologies have been a key investment area within IQT in recent years, requiring a comprehensive strategy and close coordination with investment partners like RRE.

This episode features special guest Hilary Mason, Data Scientist in Residence at Accel Partners, co-founder of hackNY.org, and founder of Fast Forward Labs, which is now part of Cloudera. Hilary joins CosmiQ’s Ryan Lewis and Nick Weir as they explore lessons learned and recommendations on executing artificial intelligence (AI) beyond research teams. The discussion explores how machine learning technologies can be applied at an enterprise level for corporations that are looking to move beyond basic prototypes and into real product development that shows results.

The final part of our three-part series featuring special guests from Azavea, Rob Emanuele, VP of Research, and Joe Morrison, Product Specialist. Join them for a discussion with CosmiQ’s Nick Weir and Jake Shermeyer as they highlight recent work exploring how to integrate some of the novel insights obtained from unique geospatial datasets, such as multi-spectral imagery and elevation information, and how these additional features can augment existing computer vision models trained on standard red, green, blue imagery and video.

Part two of a three-part series featuring special guests from Azavea, Rob Emanuele, VP of Research, and Joe Morrison, Product Specialist. In this episode, IQT’s Ryan Lewis and Azavea’s Rob and Joe dive into the workings of Azavea, a company focused on using geospatial tech and open source software to solve difficult problems across a variety of domains. Listen in to learn how businesses can be structured around open source software and what drives work in the geospatial market and beyond.

Part one of a three-part series featuring special guests from Azavea, Rob Emanuele, VP of Research, and Joe Morrison, Product Specialist. Join them for a discussion with CosmiQ’s Ryan Lewis and Nick Weir to learn how artificial intelligence (AI) and machine learning can impact the geospatial industry both technically as well as operationally, and the utility of pre-training AI models and multi-spectral satellite imagery data.

SpaceNet LLC is a non-profit organization dedicated to accelerating open source, artificial intelligence applied research for geospatial applications, specifically foundational mapping. Since its founding, the data set has been downloaded millions of times in over 80 countries. How did all of this get started? In the second of a two-part series, Ryan Lewis (CosmiQ Works), Todd Bacastow (Maxar), Joe Flasher (AWS), and Alexei Bastidas (Intel AI)  discuss how SpaceNet has evolved over the years and where they see it going in the future. 

SpaceNet LLC is a non-profit organization dedicated to accelerating open source, artificial intelligence applied research for geospatial applications, specifically foundational mapping. Since its founding, the dataset has been downloaded millions of times in more than 80 countries. How did all of this get started? In the first of a two-part series, Ryan Lewis (CosmiQ Works), Todd Bacastow (Maxar), Joe Flasher (AWS), and Alexei Bastidas (Intel AI) discuss the origins of SpaceNet and how all the different partners contribute to this initiative.  SpaceNet is a collaboration between CosmiQ Works, Maxar, Intel AI, Amazon Web Services (AWS), and Capella Space.

Dylan George from In-Q-Tel's (IQT) B.Next joins CosmiQ Works' Jake Shermeyer and Ryan Lewis to discuss how a time series of satellite imagery can be used in humanitarian response scenarios. A time series of satellite images can play a major role in these scenarios and assist practitioners in identifying areas that experience a change in population, the loss of electrical power, and/or rapid changes in land-cover. The podcast addresses two main questions: (1) How can practitioners quickly analyze these time series with open source software tools, such as CosmiQ’s CometTS tool suite; and (2) How can those tools be used to support a humanitarian response effort?  Learn how CosmiQ tackled these questions with their CometTS tool suite through the study of satellite imagery of Puerto Rico following Hurricane Maria.

Performing machine learning and analyzing geospatial data are both hard problems requiring a lot of domain expertise. These limitations have historically meant that one needs to be an expert in both to perform even the most basic analyses, making advances in AI for overhead imagery difficult to achieve. Is there anything we can do to reduce this barrier to entry, making it easier to apply machine learning methods to overhead imagery data?   Ryan Lewis and Nick Weir tackle that question as they discuss one of CosmiQ’s new project, Solaris, a new open source Python library for performing and evaluating machine learning analyses of overhead imagery. Solaris provides an easy-to-use, end-to-end analysis pipeline for AI model training, prediction, and performance assessment, along with providing pre-trained winning models from the SpaceNet® Challenges for comparison. 

Determining optimal routing paths in near real-time is at the heart of many humanitarian, civil, military, and commercial challenges. This statement is as true today as it was two years ago when the SpaceNet® partners announced the SpaceNet Challenge 3, focusing on road network detection.  On today’s pod we are joined by Ryan Lewis and Adam Van Etten to learn about SpaceNet’s upcoming fifth challenge that revisits road detection, routing, and now travel time estimation from satellite imagery. This public challenge will test participants’ ability to automatically extract road networks from satellite imagery, along with travel time estimates along all roadways, thereby permitting true optimal routing.  SpaceNet is a collaborative effort between CosmiQ Works, AWS, Maxar, and Intel AI. Learn more at www.spacenet.ai.

Consider all of the onboard analytics necessary for autonomous/semi-autonomous car operations such as lane assist. These offerings present a new attack surface for malicious actors in a myriad of industries ranging from automotive to space systems. Yet, market trends are clearly moving towards more increasing connectivity and expanding edge analytics for most of the devices we use in our daily lives. The question becomes the following: what can organizations do to increase cybersecurity of their embedded systems while these market dynamics unfold? Join us for a discussion with Red Balloon Security’s Founder and CEO, Dr. Ang Cui, and Research Scientist Joseph Pantoga, to learn about security implications (and potential solutions) of deploying artificial intelligence models to connected embedded devices. Mentioned links: https://www.blackhat.com/us-19/ https://www.redballoonsecurity.com/  https://www.escar.info/images/Datastore/2018_escar_USA/AutomotiveSandbox_Nathaniel_Boggs_Escar_USA_2018.pdf 

Peter Wang, CTO and Co-Founder of Anaconda sits down with Ryan Lewis, Adam Van Etten, and Coley Lewis to talk about open source software and its role for both data science applied research as well as product development and deployment.

Below is a link to the research paper, "SpaceNet MOVI: a Multi-View Overhead Imagery Dataset," that Nick mentioned during the episode: https://arxiv.org/abs/1903.12239.   

Adam Van Etten, Jake Shermeyer, and Ryan Lewis discuss CosmiQ's year long project studying the impact of image super resolution on object detection model performance.

Adam Van Etten, CosmiQ's Director of Research, and David Lindenbaum, formerly CosmiQ's Principal Engineer, talk about the SpaceNet 3 Challenge: Road network extract and routing from a single satellite image.

CosmiQ co-founders Ryan Lewis & David Lindenbaum talk about the motivation for startin the lab, trends in geospatial analytics, and upcoming projects.