Automatic Maasailand Boma Mapping with Deep Neural Networks

Cheng, Keli; Popescu, Ilinca; Sheets, Lincoln; Scott, Grant

doi:10.1109/igarss47720.2021.9553386

Cited by 3 publications

(9 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [2], the manual curation of the boma dataset using Google Earth and the Google Maps API is detailed. Fig.…”

Section: Tanzania Maasai Boma Datasetmentioning

confidence: 99%

“…Our data collection, as detailed in [2], is summarized as follows. Google Earth Pro was used to manually scan the study area, bounded by 3 • 58' 37" S, 36 • 09' 05" E to 4 • 31' 18" S, 36 • 43' 53" E, and bomas were marked with dropped pins in the approximated center.…”

Section: Tanzania Maasai Boma Datasetmentioning

confidence: 99%

“…That is, the system should triage the area and rank likely boma higher and lowconfidence boma lower. In [2], different DNN approaches were evaluated: the residual network, inception network, and neural architecture search (NAS). The ResNet50 [10] is a 50-layer deep vanilla residual network, while the Xception [8] is an extreme version of the inception module.…”

Section: Automated Mapping With Dnnmentioning

confidence: 99%

“…Deep neural networks (DNN) can be applied across a broad swath of area, scanning satellite imagery to provide a very cost-effective and time-saving approach for boma mapping. In [2], it was demonstrated that numerous DNN could achieve high F1 scores, over 95%, in cross-validation experiments of boma recognition. However, it is also necessary to evaluate the use of DNN in real-world applications, outside of the close-world cross-validation paradigm.…”

mentioning

confidence: 99%

“…This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ This article provides a recap from [2] for scene classification performance of various DNN for identifying RSI chips with boma. We then further develop the process to leverage the top-performing DNN for real-world mapping over a large swath of earth in Tanzania.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Analysis of Deep Learning Techniques for Maasai Boma Mapping in Tanzania

Cheng

Popescu

Sheets

et al. 2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

Underdeveloped countries in sub-Saharan Africa often contain cultural subpopulations that are underserved in regard to health and education. This perpetuates the health challenges of the country as a whole, and it is therefore of interest to be able to automatically map the subpopulation for the health services delivery. International nonprofit health organizations have often taken the lead in these efforts, providing humanitarian aid (e.g., clean water and food) as well as health care. This is necessary, as the ethnic subpopulations are not well integrated into the society and the existing health care systems. In this study, we explicitly explore the Maasailand of Tanzania, to evaluate the use of deep neural networks (DNN) to aid in the automatic visual analysis of remote sensing data to geolocate Maasai boma structures. We investigate the performance of four state-of-the-art DNN as classifiers of boma presence within high-resolution imagery; all showing over 95% F1 score performance. Additionally, we scan over 3900 km 2 of high-resolution imagery, combining a ProxylessNAS with broad area aggregation and mapping techniques and demonstrate the discovery of hundreds of boma, many that were not discovered by human analysts performing visual scans. The trained Proxyless-NAS model generates a classified vector response field (CVRF). The CVRF is aggregated by a mode-seeking algorithm to detect potential locations of boma structures within the study area. The model detected numerous human false negatives (HFNs) and achieved 94.022% TPR and 95.395% F1 score using an aggregation aperture of 250 m within a 76.620 square kilometers area of interest.

show abstract

“…In [2], the manual curation of the boma dataset using Google Earth and the Google Maps API is detailed. Fig.…”

Section: Tanzania Maasai Boma Datasetmentioning

confidence: 99%

Section: Tanzania Maasai Boma Datasetmentioning

confidence: 99%

Section: Automated Mapping With Dnnmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Analysis of Deep Learning Techniques for Maasai Boma Mapping in Tanzania

Cheng

Popescu

Sheets

et al. 2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

Evaluation of Sentinel-2 Data for Automatic Maasai Boma Mapping

Cheng

Bajkowski

Scott

2021

2021 IEEE Applied Imagery Pattern Recognition Workshop (AIPR)

View full text Add to dashboard Cite

Deep Seasonal Network for Remote Sensing Imagery Classification of Multi-Temporal Sentinel-2 Data

Cheng,

Scott

2023

Remote Sensing

Self Cite

View full text Add to dashboard Cite

As a medium-resolution multi-temporal data source, Sentinel-2 data has the potential to match the performance of using very-high-resolution (VHR) images in deep learning applications. To fully leverage the multi-temporal nature of Sentinel-2 data, we introduce the Deep Seasonal Network (DeepSN). This composite architecture combines a pre-trained deep convolutional neural network (DCNN) for visual feature extraction with a long short-term memory (LSTM) model to capture temporal information and make classification predictions. We evaluate the effectiveness of DeepSN on a Maasai Boma classification task in the Tanzania region. The DeepSN takes a sequence of four seasonal data, each spanning three months, for Boma prediction. Through cross-season validation experiments, we compare various advanced DCNNs and select EfficientNet as the backbone for DeepSN, as it performs the best. DeepSN with an EfficientNet backbone achieves a significant 19% improvement in the F1 score compared to plain EfficientNet for the Boma classification task. This work introduces a versatile composite architecture capable of handling multi-temporal data efficiently, providing flexibility in choosing the most suitable feature extraction backbone. The performance of DeepSN demonstrates the viability of utilizing medium-resolution multi-temporal data instead of high-resolution images for diverse tasks.

show abstract

Automatic Maasailand Boma Mapping with Deep Neural Networks

Cited by 3 publications

References 13 publications

Analysis of Deep Learning Techniques for Maasai Boma Mapping in Tanzania

Analysis of Deep Learning Techniques for Maasai Boma Mapping in Tanzania

Evaluation of Sentinel-2 Data for Automatic Maasai Boma Mapping

Deep Seasonal Network for Remote Sensing Imagery Classification of Multi-Temporal Sentinel-2 Data

Contact Info

Product

Resources

About