Optimizing Medical Response Time through Deep Learning Based Accident Detection
Algorithms and automatic surveillance systems are increasingly being used in regulation enforcement across many countries. Unfortunately, this technological leap hasn’t yet benefited the safety of road users, resulting in many unnecessary casualties due to too long response time from medical services even in developed urban regions. We analyzed the different elements of what constitutes the time it takes to arrive on an accident location to reduce the delay using modern technology.Our tool, an easy-to-use website based on our own developped AI, provides a real-time traffic analysis from custom live footage with a very high rate of accident detection. On our test sets, the accident detection was close to 100% with some false positives. These preliminary results highlight the potential for the tool to hasten medical care response time and reduce the intervention time significantly, increasing the chances of survival of car accident victims.
Authors
- Thomas M. Hemmerling (ORCID: https://orcid.org/0000-0002-1447-0703)
- Sean Jeffries
- Nicolas François (ORCID: https://orcid.org/0000-0003-1472-5945)
- Franck Staudt
- Raphaël Olexa
- Georgiy Danylenko
Institutions
- Xerox (France) (FR)
- Université franco-allemande (DE)
- McGill University (CA)
Publication Details
- Published
- 2025-08-07
- DOI
- https://doi.org/10.1109/acdsa65407.2025.11166298
- Primary Topic
- IoT and GPS-based Vehicle Safety Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00