Current Result Document :
| 한글제목(Korean Title) |
Federated Compact Transformers for Intelligent Fire Detection System in Smart City |
| 영문제목(English Title) |
Federated Compact Transformers for Intelligent Fire Detection System in Smart City |
| 저자(Author) |
Chu Myaet Thwal
Ye Lin Tun
Choong Seon Hong
|
| 원문수록처(Citation) |
VOL 48 NO. 02 PP. 0269 ~ 0271 (2021. 12) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Large, extended-day wildfires and fire incidents in high-population areas may lead to major damages and emergency situations. Thus, early-stage fire detection with high sensitivity and accuracy becomes a challenging task in both indoor and outdoor cases. With the rise of artificial intelligence and computer vision, fire detection through surveillance cameras has proven to be an innovative solution in the smart city domain. However, the efficiency in model performance and communication cost is critical in such applications. In this work, we propose a federated learning based fire detection system where distributed edge servers collaboratively participate in training a global model without revealing their private data for security and privacy concerns. We implement lightweight vision transformer models and leverage the use of decentralized training data. By applying the proposed intelligent fire detection scheme in the smart city domain, it can be expected to reduce delays in emergency operations, thus providing greater and smarter facilities for the citizens. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
