In this blog we explore the efficacy and drawbacks of adversarial camouflage in an overhead imagery context. Note: This blog is a summary of our ArXiv paper. 1. Executive Summary While a number of recent papers have demonstrated the ability to reliably fool deep learning classifiers and object detectors with adversarial patches (see…

Foreword: This article first appeared in the IQT blog, reproduced here with permission. Executive Summary With the increasing availability of overhead imagery, machine learning is becoming a critical tool for analyzing this imagery. Advanced computer vision object detection techniques have demonstrated great success in identifying objects of interest such as…

When standard navigation techniques (i.e. Google Maps) are down and time is of the essence, how might one evacuate from (or position aid to) a hazardous locale? In this post we explore one option, which we call Project Shackleton. As we’ve noted before, in a disaster scenario (be it an…

A lightweight codebase for building a local tile server for Bokeh applications — In our last post we discussed methods for vehicle localization, congestion detection, and optimized routing with the open-source Diet Hadrade repository. Diet Hadrade allows users to experiment with static images within a Jupyter notebook. As we look to build upon this functionality, one critical piece is tiling imagery rather than…

Foreword: This blog first appeared on the IQT blog, and is the product of Adam Van Etten and Daniel Hogan. Reproduced here with permission. 1. Executive Summary Much ink has been spilled on systemic causes of hidden technical debt of machine learning systems (Sculley et al, 2015). Such issues include: “hidden feedback loops…

Foreword: This blog first appeared on the IQT blog as part of an ongoing IQT Labs research project exploring synthetic data and satellite imagery. Reproduced here with permission. 1. Introduction The RarePlanes satellite imagery dataset is rich enough to enable copious machine learning and object detection studies, particularly when coupled…

A lightweight machine learning codebase for aiding HADR (humanitarian assistance and disaster relief) Humanitarian crises, both natural and man-made, show no sign of abating in the near term given ongoing changes to the planet and society. Yet ever improving machine learning techniques hold the possibility of blunting some of the…

Announcing the release of YOLTv5 Preface: This blog is part of a series describing the work done at Geodesic Labs. The YOLO family of deep learning object detection models has enjoyed remarkable longevity, with new versions still under active development seven years after the original paper release. We have actively…

Announcing Geodesic Labs — Geodesic Labs was founded a couple years ago with the premise of leveraging remote sensing imagery and machine learning techniques to aid in situational awareness and decision optimization in austere environments. Specifically, the rapid extraction of both vehicles and road networks from imagery allows a host of interesting problems to…