SaRNet: A Dataset for Deep Learning Assisted Search and Rescue with Satellite Imagery

Thoreau, Michael; Wilson, Frazer

doi:10.1109/ispa52656.2021.9552103

Cited by 3 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A study published in 2021 SaRNet: A Dataset for Deep Learning Assisted Search and Rescue with Satellite Imagery (Thoreau & Wilson, 2021), looks at high resolution satellite imagery with dramatically improved resolution for humanitarian relief and Search and Rescue (SAR). The proposal uses a novel remote sensing object detection dataset for deep learning assisted SAR.…”

Section: Websites and Applicationsmentioning

confidence: 99%

Untitled

2024

JSAR

View full text Add to dashboard Cite

Readers will have noticed, in the previous pages, that this issue contains for the first time a list of our corporate partnerships and affiliations. In the last issue, we announced our partnership with the Institute of Civil Protection and Emergency Management (ICPEM) and published a guest editorial from their President. This resulted in two benefits for this journal. The first is that ICPEM are sponsoring the production of digital object identifiers DOIs for all articles published in our Journal. This has meant that papers published here are now referenced in all major international libraries and cross-referenced for their discussion in in other publications. The result of that is that many of our contributing authors have seen their research rankings improved, as well as an increased influence of Search and Rescue as a scientific discipline.The second beneficial result was that a number of other relevant organisations contacted us to discuss alliances and affiliations with them and I'm proud to include a page in this issue sharing the details of such an illustrious collection of Search and Rescue Organisations. There are links below each logo, so please do visit them, and if appropriate, consider joining them as a member. This issue is split into three parts. The first section is the typical contributions original research. We then have two letters to the editor, and two submissions to the annual Syrotuck Symposium held in Iceland in 2023.Discussions of Artificial Intelligence appear in the articles in this issue, and did in the previous issue. In addition it is in evidence on the cover, where artificial intelligence was used to create a futuristic Search and Rescue technician.Artificial Intelligence often suffers from a bad press internationally with people (perhaps justifiably) concerned that without adequate controls there may be risks to humanity. In addition, people are reasonably concerned that artificial intelligence may replace them at work and that our social and economic systems have not developed adequate protection for workers replaced by artificial intelligence. Although these concerns are legitimate, I prefer to see artificial intelligence as an opportunity rather than a threatespecially in the realm of search and rescue.The ability to tap into many centuries of knowledge and experience in seconds when planning conducting or analysing data relating to a search has enormous potential for our sector. Admittedly, some of the same concerns may exist, in that we may need fewer search planners if artificial intelligence becomes reliable enough, but this is the same with all advances in technology. I would argue that in our sector where it

show abstract

Section: Websites and Applicationsmentioning

confidence: 99%

Untitled

2024

JSAR

View full text Add to dashboard Cite

show abstract

“…Imagery (Thoreau & Wilson, 2021), looks at high resolution satellite imagery with dramatically improved resolution for humanitarian relief and Search and Rescue (SAR). The proposal uses a novel remote sensing object detection dataset for deep learning assisted SAR.…”

Section: A Study Published In 2021 Sarnet: a Dataset For Deep Learnin...mentioning

confidence: 99%

The Search Intelligence Process Using Artificial Intelligence

Young,

Kurfess,

Bloom

2024

JSAR

View full text Add to dashboard Cite

Intelligence is a process by which certain types of information/data are aquired or requested by those directing a missing person incident. The process continues by rallying teams of resources to collect, process and exploit, analyze and produce useful information that can then be disseminated and integrated into actionable plans. Traditionally the collection, processing, exploitation, analysis, and production has been done by hand on paper and sorted by experienced search leaders and training of the human brain. However, what if this work could be done by computers and artificial intelligence? This is the hypothesis we are exploring through a privately funded university effort. The “Artificial Intelligence for Search and Rescue” is a project with goals to use AI and related computational methods and tools to support Search and Rescue (SAR) missions. This project is a collaboration with California Polytechnic State University (Cal Poly) Computer Science and Software Engineer Department Dr. Franz Kurfess, undergraduate and graduate students in alliance with alumnus Gary Bloom and Christopher Young both experienced search and rescue incident managers. The project began in the summer of 2021 through privately funded research, senior projects, and class projects. This paper is Part 1 and will describe the initial work included the collecting and consolidating the pertinent mechanisms for storing and processing data and the research of the potential artificial intelligence options. Part 2 will describe the final product after lab/field testing and refinements. KEY WORDS: Intelligence Gathering, Missing Person, Interviewing.

show abstract