An Integrated YOLO and VLM System for Fire Detection in Enclosed Environments

Jongeun Kim, Yejin Lee, Dongsik Yoon, Chansung Jung, Gunhee Lee
Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops, PMLR 296:151-162, 2025.

Abstract

While YOLO models show promise in car fire detection, they remain insufficient for real-world deployment in confined parking environments due to dataset limitations, evaluation gaps, and deployment constraints. We first fine-tune YOLO on a fire/smoke-augmented dataset, but analysis reveals its struggles with ambiguous fire-smoke boundaries, leading to false predictions. To address this, we propose a real-time end-to-end framework integrating YOLOv8s with Florence2 VLM, combining object detection with contextual reasoning. While YOLOv8s with VLM improves detection reliability, challenges are still ongoing. Our findings highlight YOLO’s limitations in fire detection and the need for a more adaptive, environment-aware approach.

Cite this Paper

BibTeX
@InProceedings{pmlr-v296-kim25a, title = {An Integrated YOLO and VLM System for Fire Detection in Enclosed Environments}, author = {Kim, Jongeun and Lee, Yejin and Yoon, Dongsik and Jung, Chansung and Lee, Gunhee}, booktitle = {Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops}, pages = {151--162}, year = {2025}, editor = {Blaas, Arno and D’Costa, Priya and Feng, Fan and Kriegler, Andreas and Mason, Ian and Pan, Zhaoying and Uelwer, Tobias and Williams, Jennifer and Xie, Yubin and Yang, Rui}, volume = {296}, series = {Proceedings of Machine Learning Research}, month = {28 Apr}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v296/main/assets/kim25a/kim25a.pdf}, url = {https://proceedings.mlr.press/v296/kim25a.html}, abstract = {While YOLO models show promise in car fire detection, they remain insufficient for real-world deployment in confined parking environments due to dataset limitations, evaluation gaps, and deployment constraints. We first fine-tune YOLO on a fire/smoke-augmented dataset, but analysis reveals its struggles with ambiguous fire-smoke boundaries, leading to false predictions. To address this, we propose a real-time end-to-end framework integrating YOLOv8s with Florence2 VLM, combining object detection with contextual reasoning. While YOLOv8s with VLM improves detection reliability, challenges are still ongoing. Our findings highlight YOLO’s limitations in fire detection and the need for a more adaptive, environment-aware approach.} }
Endnote
%0 Conference Paper %T An Integrated YOLO and VLM System for Fire Detection in Enclosed Environments %A Jongeun Kim %A Yejin Lee %A Dongsik Yoon %A Chansung Jung %A Gunhee Lee %B Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops %C Proceedings of Machine Learning Research %D 2025 %E Arno Blaas %E Priya D’Costa %E Fan Feng %E Andreas Kriegler %E Ian Mason %E Zhaoying Pan %E Tobias Uelwer %E Jennifer Williams %E Yubin Xie %E Rui Yang %F pmlr-v296-kim25a %I PMLR %P 151--162 %U https://proceedings.mlr.press/v296/kim25a.html %V 296 %X While YOLO models show promise in car fire detection, they remain insufficient for real-world deployment in confined parking environments due to dataset limitations, evaluation gaps, and deployment constraints. We first fine-tune YOLO on a fire/smoke-augmented dataset, but analysis reveals its struggles with ambiguous fire-smoke boundaries, leading to false predictions. To address this, we propose a real-time end-to-end framework integrating YOLOv8s with Florence2 VLM, combining object detection with contextual reasoning. While YOLOv8s with VLM improves detection reliability, challenges are still ongoing. Our findings highlight YOLO’s limitations in fire detection and the need for a more adaptive, environment-aware approach.
APA
Kim, J., Lee, Y., Yoon, D., Jung, C. & Lee, G.. (2025). An Integrated YOLO and VLM System for Fire Detection in Enclosed Environments. Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops, in Proceedings of Machine Learning Research 296:151-162 Available from https://proceedings.mlr.press/v296/kim25a.html.

Related Material