Adam Van Etten

Adam Van Etten

Machine learning is increasingly critical for analysis of the ever-growing corpora of overhead imagery. Advanced computer vision object detection techniques have demonstrated great success in identifying objects of interest such as ships, automobiles, and aircraft from satellite and drone imagery. Yet relying on computer vision opens up significant vulnerabilities, namely, the susceptibility of object detection algorithms to adversarial attacks. In this paper we explore the efficacy and drawbacks of adversarial camouflage in an overhead imagery context. While a number of recent papers have demonstrated the ability to reliably fool deep learning classifiers and object detectors with adversarial patches, most of this work has been performed on relatively uniform datasets and only a single class of objects. In this work we utilize the VisDrone dataset, which has a large range of perspectives and object sizes. We explore four different object classes: bus, car, truck, van. We build a library of 24 adversarial patches to disguise these objects, and introduce a patch translucency variable to our patches. The translucency (or alpha value) of the patches is highly correlated to their efficacy. Further, we show that while adversarial patches may fool object detectors, the presence of such patches is often easily uncovered, with patches on average 24% more detectable than the objects the patches were meant to hide. This raises the question of whether such patches truly constitute camouflage. Source code is available at https://github.com/IQTLabs/camolo.

Adam Van Etten, Daniel Hogan

Building footprints provide a useful proxy for a great many humanitarian applications. For example, building footprints are useful for high fidelity population estimates, and quantifying population statistics is fundamental to ~1/4 of the United Nations Sustainable Development Goals Indicators. In this paper we (the SpaceNet Partners) discuss efforts to develop techniques for precise building footprint localization, tracking, and change detection via the SpaceNet Multi-Temporal Urban Development Challenge (also known as SpaceNet 7). In this NeurIPS 2020 competition, participants were asked identify and track buildings in satellite imagery time series collected over rapidly urbanizing areas. The competition centered around a brand new open source dataset of Planet Labs satellite imagery mosaics at 4m resolution, which includes 24 images (one per month) covering ~100 unique geographies. Tracking individual buildings at this resolution is quite challenging, yet the winning participants demonstrated impressive performance with the newly developed SpaceNet Change and Object Tracking (SCOT) metric. This paper details the top-5 winning approaches, as well as analysis of results that yielded a handful of interesting anecdotes such as decreasing performance with latitude.

Adam Van Etten, Daniel Hogan, Jesus Martinez-Manso et al.

Satellite imagery analytics have numerous human development and disaster response applications, particularly when time series methods are involved. For example, quantifying population statistics is fundamental to 67 of the 231 United Nations Sustainable Development Goals Indicators, but the World Bank estimates that over 100 countries currently lack effective Civil Registration systems. To help address this deficit and develop novel computer vision methods for time series data, we present the Multi-Temporal Urban Development SpaceNet (MUDS, also known as SpaceNet 7) dataset. This open source dataset consists of medium resolution (4.0m) satellite imagery mosaics, which includes 24 images (one per month) covering >100 unique geographies, and comprises >40,000 km2 of imagery and exhaustive polygon labels of building footprints therein, totaling over 11M individual annotations. Each building is assigned a unique identifier (i.e. address), which permits tracking of individual objects over time. Label fidelity exceeds image resolution; this "omniscient labeling" is a unique feature of the dataset, and enables surprisingly precise algorithmic models to be crafted. We demonstrate methods to track building footprint construction (or demolition) over time, thereby directly assessing urbanization. Performance is measured with the newly developed SpaceNet Change and Object Tracking (SCOT) metric, which quantifies both object tracking as well as change detection. We demonstrate that despite the moderate resolution of the data, we are able to track individual building identifiers over time. This task has broad implications for disaster preparedness, the environment, infrastructure development, and epidemic prevention.

Jacob Shermeyer, Thomas Hossler, Adam Van Etten et al.

RarePlanes is a unique open-source machine learning dataset that incorporates both real and synthetically generated satellite imagery. The RarePlanes dataset specifically focuses on the value of synthetic data to aid computer vision algorithms in their ability to automatically detect aircraft and their attributes in satellite imagery. Although other synthetic/real combination datasets exist, RarePlanes is the largest openly-available very-high resolution dataset built to test the value of synthetic data from an overhead perspective. Previous research has shown that synthetic data can reduce the amount of real training data needed and potentially improve performance for many tasks in the computer vision domain. The real portion of the dataset consists of 253 Maxar WorldView-3 satellite scenes spanning 112 locations and 2,142 km^2 with 14,700 hand-annotated aircraft. The accompanying synthetic dataset is generated via AI.Reverie's simulation platform and features 50,000 synthetic satellite images simulating a total area of 9331.2 km^2 with ~630,000 aircraft annotations. Both the real and synthetically generated aircraft feature 10 fine grain attributes including: aircraft length, wingspan, wing-shape, wing-position, wingspan class, propulsion, number of engines, number of vertical-stabilizers, presence of canards, and aircraft role. Finally, we conduct extensive experiments to evaluate the real and synthetic datasets and compare performances. By doing so, we show the value of synthetic data for the task of detecting and classifying aircraft from an overhead perspective.

Adam Van Etten

Automated road network extraction from remote sensing imagery remains a significant challenge despite its importance in a broad array of applications. To this end, we leverage recent open source advances and the high quality SpaceNet dataset to explore road network extraction at scale, an approach we call City-scale Road Extraction from Satellite Imagery (CRESI). Specifically, we create an algorithm to extract road networks directly from imagery over city-scale regions, which can subsequently be used for routing purposes. We quantify the performance of our algorithm with the APLS and TOPO graph-theoretic metrics over a diverse 608 square kilometer test area covering four cities. We find an aggregate score of APLS = 0.73, and a TOPO score of 0.58 (a significant improvement over existing methods). Inference speed is 160 square kilometers per hour on modest hardware. Finally, we demonstrate that one can use the extracted road network for any number of applications, such as optimized routing.

Nicholas Weir, David Lindenbaum, Alexei Bastidas et al.

Detection and segmentation of objects in overheard imagery is a challenging task. The variable density, random orientation, small size, and instance-to-instance heterogeneity of objects in overhead imagery calls for approaches distinct from existing models designed for natural scene datasets. Though new overhead imagery datasets are being developed, they almost universally comprise a single view taken from directly overhead ("at nadir"), failing to address a critical variable: look angle. By contrast, views vary in real-world overhead imagery, particularly in dynamic scenarios such as natural disasters where first looks are often over 40 degrees off-nadir. This represents an important challenge to computer vision methods, as changing view angle adds distortions, alters resolution, and changes lighting. At present, the impact of these perturbations for algorithmic detection and segmentation of objects is untested. To address this problem, we present an open source Multi-View Overhead Imagery dataset, termed SpaceNet MVOI, with 27 unique looks from a broad range of viewing angles (-32.5 degrees to 54.0 degrees). Each of these images cover the same 665 square km geographic extent and are annotated with 126,747 building footprint labels, enabling direct assessment of the impact of viewpoint perturbation on model performance. We benchmark multiple leading segmentation and object detection models on: (1) building detection, (2) generalization to unseen viewing angles and resolutions, and (3) sensitivity of building footprint extraction to changes in resolution. We find that state of the art segmentation and object detection models struggle to identify buildings in off-nadir imagery and generalize poorly to unseen views, presenting an important benchmark to explore the broadly relevant challenge of detecting small, heterogeneous target objects in visually dynamic contexts.

Adam Van Etten

Detection of small objects in large swaths of imagery is one of the primary problems in satellite imagery analytics. While object detection in ground-based imagery has benefited from research into new deep learning approaches, transitioning such technology to overhead imagery is nontrivial. Among the challenges is the sheer number of pixels and geographic extent per image: a single DigitalGlobe satellite image encompasses >64 km2 and over 250 million pixels. Another challenge is that objects of interest are minuscule (often only ~10 pixels in extent), which complicates traditional computer vision techniques. To address these issues, we propose a pipeline (You Only Look Twice, or YOLT) that evaluates satellite images of arbitrary size at a rate of >0.5 km2/s. The proposed approach can rapidly detect objects of vastly different scales with relatively little training data over multiple sensors. We evaluate large test images at native resolution, and yield scores of F1 > 0.8 for vehicle localization. We further explore resolution and object size requirements by systematically testing the pipeline at decreasing resolution, and conclude that objects only ~5 pixels in size can still be localized with high confidence. Code is available at https://github.com/CosmiQ/yolt.

Adam Van Etten

We explore methods to detect automobiles in Planet imagery and build a large scale vector field for moving objects. Planet operates two distinct constellations: high-resolution SkySat satellites as well as medium-resolution SuperDove satellites. We show that both static and moving cars can be identified reliably in high-resolution SkySat imagery. We are able to estimate the speed and heading of moving vehicles by leveraging the inter-band displacement (or "rainbow" effect) of moving objects. Identifying cars and trucks in medium-resolution SuperDove imagery is far more difficult, though a similar rainbow effect is observed in these satellites and enables moving vehicles to be detected and vectorized. The frequent revisit of Planet satellites enables the categorization of automobile and truck activity patterns over broad areas of interest and lengthy timeframes.

Jacob Shermeyer, Daniel Hogan, Jason Brown et al.

Within the remote sensing domain, a diverse set of acquisition modalities exist, each with their own unique strengths and weaknesses. Yet, most of the current literature and open datasets only deal with electro-optical (optical) data for different detection and segmentation tasks at high spatial resolutions. optical data is often the preferred choice for geospatial applications, but requires clear skies and little cloud cover to work well. Conversely, Synthetic Aperture Radar (SAR) sensors have the unique capability to penetrate clouds and collect during all weather, day and night conditions. Consequently, SAR data are particularly valuable in the quest to aid disaster response, when weather and cloud cover can obstruct traditional optical sensors. Despite all of these advantages, there is little open data available to researchers to explore the effectiveness of SAR for such applications, particularly at very-high spatial resolutions, i.e. <1m Ground Sample Distance (GSD). To address this problem, we present an open Multi-Sensor All Weather Mapping (MSAW) dataset and challenge, which features two collection modalities (both SAR and optical). The dataset and challenge focus on mapping and building footprint extraction using a combination of these data sources. MSAW covers 120 km^2 over multiple overlapping collects and is annotated with over 48,000 unique building footprints labels, enabling the creation and evaluation of mapping algorithms for multi-modal data. We present a baseline and benchmark for building footprint extraction with SAR data and find that state-of-the-art segmentation models pre-trained on optical data, and then trained on SAR (F1 score of 0.21) outperform those trained on SAR data alone (F1 score of 0.135).

Adam Van Etten, Jacob Shermeyer, Daniel Hogan et al.

Identification of road networks and optimal routes directly from remote sensing is of critical importance to a broad array of humanitarian and commercial applications. Yet while identification of road pixels has been attempted before, estimation of route travel times from overhead imagery remains a novel problem, particularly for off-nadir overhead imagery. To this end, we extract road networks with travel time estimates from the SpaceNet MVOI dataset. Utilizing the CRESIv2 framework, we demonstrate the ability to extract road networks in various observation angles and quantify performance at 27 unique nadir angles with the graph-theoretic APLS_length and APLS_time metrics. A minimal gap of 0.03 between APLS_length and APLS_time scores indicates that our approach yields speed limits and travel times with very high fidelity. We also explore the utility of incorporating all available angles during model training, and find a peak score of APLS_time = 0.56. The combined model exhibits greatly improved robustness over angle-specific models, despite the very different appearance of road networks at extremely oblique off-nadir angles versus images captured from directly overhead.

Adam Van Etten

Automated road network extraction from remote sensing imagery remains a significant challenge despite its importance in a broad array of applications. To this end, we explore road network extraction at scale with inference of semantic features of the graph, identifying speed limits and route travel times for each roadway. We call this approach City-Scale Road Extraction from Satellite Imagery v2 (CRESIv2), Including estimates for travel time permits true optimal routing (rather than just the shortest geographic distance), which is not possible with existing remote sensing imagery based methods. We evaluate our method using two sources of labels (OpenStreetMap, and those from the SpaceNet dataset), and find that models both trained and tested on SpaceNet labels outperform OpenStreetMap labels by greater than 60%. We quantify the performance of our algorithm with the Average Path Length Similarity (APLS) and map topology (TOPO) graph-theoretic metrics over a diverse test area covering four cities in the SpaceNet dataset. For a traditional edge weight of geometric distance, we find an aggregate of 5% improvement over existing methods for SpaceNet data. We also test our algorithm on Google satellite imagery with OpenStreetMap labels, and find a 23% improvement over previous work. Metric scores decrease by only 4% on large graphs when using travel time rather than geometric distance for edge weights, indicating that optimizing routing for travel time is feasible with this approach.

Jacob Shermeyer, Adam Van Etten

We explore the application of super-resolution techniques to satellite imagery, and the effects of these techniques on object detection algorithm performance. Specifically, we enhance satellite imagery beyond its native resolution, and test if we can identify various types of vehicles, planes, and boats with greater accuracy than native resolution. Using the Very Deep Super-Resolution (VDSR) framework and a custom Random Forest Super-Resolution (RFSR) framework we generate enhancement levels of 2x, 4x, and 8x over five distinct resolutions ranging from 30 cm to 4.8 meters. Using both native and super-resolved data, we then train several custom detection models using the SIMRDWN object detection framework. SIMRDWN combines a number of popular object detection algorithms (e.g. SSD, YOLO) into a unified framework that is designed to rapidly detect objects in large satellite images. This approach allows us to quantify the effects of super-resolution techniques on object detection performance across multiple classes and resolutions. We also quantify the performance of object detection as a function of native resolution and object pixel size. For our test set we note that performance degrades from mean average precision (mAP) = 0.53 at 30 cm resolution, down to mAP = 0.11 at 4.8 m resolution. Super-resolving native 30 cm imagery to 15 cm yields the greatest benefit; a 13-36% improvement in mAP. Super-resolution is less beneficial at coarser resolutions, though still provides a small improvement in performance.

Adam Van Etten

Detecting small objects over large areas remains a significant challenge in satellite imagery analytics. Among the challenges is the sheer number of pixels and geographical extent per image: a single DigitalGlobe satellite image encompasses over 64 km2 and over 250 million pixels. Another challenge is that objects of interest are often minuscule (~pixels in extent even for the highest resolution imagery), which complicates traditional computer vision techniques. To address these issues, we propose a pipeline (SIMRDWN) that evaluates satellite images of arbitrarily large size at native resolution at a rate of > 0.2 km2/s. Building upon the tensorflow object detection API paper, this pipeline offers a unified approach to multiple object detection frameworks that can run inference on images of arbitrary size. The SIMRDWN pipeline includes a modified version of YOLO (known as YOLT), along with the models of the tensorflow object detection API: SSD, Faster R-CNN, and R-FCN. The proposed approach allows comparison of the performance of these four frameworks, and can rapidly detect objects of vastly different scales with relatively little training data over multiple sensors. For objects of very different scales (e.g. airplanes versus airports) we find that using two different detectors at different scales is very effective with negligible runtime cost.We evaluate large test images at native resolution and find mAP scores of 0.2 to 0.8 for vehicle localization, with the YOLT architecture achieving both the highest mAP and fastest inference speed.

Adam Van Etten, Dave Lindenbaum, Todd M. Bacastow

Foundational mapping remains a challenge in many parts of the world, particularly in dynamic scenarios such as natural disasters when timely updates are critical. Updating maps is currently a highly manual process requiring a large number of human labelers to either create features or rigorously validate automated outputs. We propose that the frequent revisits of earth imaging satellite constellations may accelerate existing efforts to quickly update foundational maps when combined with advanced machine learning techniques. Accordingly, the SpaceNet partners (CosmiQ Works, Radiant Solutions, and NVIDIA), released a large corpus of labeled satellite imagery on Amazon Web Services (AWS) called SpaceNet. The SpaceNet partners also launched a series of public prize competitions to encourage improvement of remote sensing machine learning algorithms. The first two of these competitions focused on automated building footprint extraction, and the most recent challenge focused on road network extraction. In this paper we discuss the SpaceNet imagery, labels, evaluation metrics, prize challenge results to date, and future plans for the SpaceNet challenge series.