Abstract
As satellite imagery collections continue to grow at an astonishing rate, so is the demand for automated and scalable object detection and segmentation. Scaling computational activities demand models that generalize well across various challenges that can hamper progress, including diverse imaging and geographic conditions, sampling bias in training data, manual ground truth collection, tooling for model reuse and accountability assessment, and poor model training strategies. A great deal of progress has been made on these challenges. We contribute to the improvement through further development of ReSFlow, a workflow that breaks the problem of model generalization into a collection of specialized exploitations. ReSFlow partitions imagery collections into homogeneous buckets equipped with exploitation models trained to perform well under each bucket.s specific context. Essentially, ReSFlow aims for generalization through stratification. Therefore, within a bucket, exploitation is a homogeneous process that mitigates heterogeneity challenges, including the number of training data and data biases that can occur over varied conditions. Furthermore, custom model architectures and rich training strategies effective for within-bucket conditions can be developed. Meanwhile, across buckets, performance metrics support systematic views of the workflow leading to optimal data labeling allocations and indications that further specialization is warranted. Herein, we discuss the formation of models during the framework's “Offline Initialization” stage. Lastly, we exploit the inherent parallelism due to bucketing to introduce model reuse and demonstrate efficacy by reducing an 89-day manual data labeling cost to zero-days in a new area of interest.
Highlights
The field of remote sensing (RS) has recently experienced a surge of research breakthroughs to long-standing application problems
Several key factors have been instrumental in providing these advances, including an abundance of data produced by better sensing technologies, the availability of high-performance computing resources, and new advances in machine learning (ML) to enable data-driven analytics
Case Study: Building Segmentation We describe both quantitative and qualitative results of the proposed ReSFlow stratification of data acquired under different imaging conditions across different geographic locations
Summary
The field of remote sensing (RS) has recently experienced a surge of research breakthroughs to long-standing application problems. Despite the recent significant progress from machine learning and computer vision in commercial applications, several characteristics of satellite imagery present unique challenges which currently limit the functionality and impact of existing methods. We summarize these challenges in the following paragraphs. Heterogeneity in remote sensing instruments: Global satellite image datasets contain high-resolution imagery collected from multiple sensing instruments at different spatial, spectral, and temporal resolutions (see Figure 1) Varying characteristics within these sensors, which include ground sampling distance and noise from sensor interference, continue to limit large-scale object detection and segmentation with machine learning and satellite imagery [5], [6]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.