Distilling Vision-Language Models for On-Device Fire Understanding

• URL: http://arxiv.org/abs/2609.05782v1
• Date: Sat, 05 Sep 2026 00:33:01 GMT
• Title: Distilling Vision-Language Models for On-Device Fire Understanding
• Abstract summary: Vision-language models (VLMs) offer a promising alternative to conventional fire detection systems.<n>Their large model size makes deployment on embedded fire sensors impractical.<n>We develop a teacher-student knowledge distillation framework in which large VLMs fine-tuned for fire understanding can be distilled into lightweight students.
• Score: 2.708990023212152
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Vision-language models (VLMs) offer a promising alternative to conventional fire detection systems by reasoning about the semantic context of a scene and thus reducing false alarms, yet their large model size makes deployment on embedded fire sensors impractical. In this paper, we study how domain-specialized VLMs can be compressed for fully on-device deployment without losing the safety-critical behavior required for fire detection. We develop a teacher-student knowledge distillation framework in which large VLMs fine-tuned for fire understanding can be distilled into lightweight students. Experiments across multiple VLM families and model scales show that compact students preserve most of their teachers' fire-understanding capability. We further deploy the distilled models on our commercial Detectium fire detection sensor and jointly evaluate reasoning accuracy, latency, and memory usage. The results show that compression and deployment affect not only accuracy but also model failure modes, with Qwen2.5-0.5B providing the strongest overall deployment trade-off. Our findings provide broader guidance for deploying domain-specialized VLMs in resource-constrained, safety-critical settings.

Related papers

• GazeCLIP: Gaze-Guided CLIP with Adaptive-Enhanced Fine-Grained Language Prompt for Deepfake Attribution and Detection [80.12497948980378]
  Current deepfake attribution or deepfake detection works tend to exhibit poor generalization to novel generative methods.<n>We propose a novel gaze-guided CLIP with adaptive-enhanced fine-grained language prompts for fine-grained deepfake attribution and detection (DFAD)<n>We conduct a novel and fine-grained benchmark to evaluate the DFAD performance of networks on novel generators like diffusion and flow models.<n>We introduce a gaze-aware model based on CLIP, which is devised to enhance the generalization to unseen face forgery attacks.
  arXiv  Detail & Related papers  (2026-03-31T05:59:59Z)
• PyroFocus: A Deep Learning Approach to Real-Time Wildfire Detection in Multispectral Remote Sensing Imagery [0.0]
  Rapid and accurate wildfire detection is crucial for emergency response and environmental management.<n>In airborne and spaceborne missions, real-time algorithms must distinguish between no fire, active fire, and post-fire conditions.<n>We introduce PyroFocus, a two-stage pipeline that performs fire classification followed by fire radiative power (FRP) regression or segmentation to reduce inference time and computational cost for onboard deployment.
  arXiv  Detail & Related papers  (2025-12-02T21:59:45Z)
• V-Attack: Targeting Disentangled Value Features for Controllable Adversarial Attacks on LVLMs [66.81402538540458]
  We propose V-Attack, a novel method for precise local semantic attacks.<n>V-Attack improves the attack success rate by an average of 36% over state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-11-25T11:51:17Z)
• Phantom Menace: Exploring and Enhancing the Robustness of VLA Models against Physical Sensor Attacks [17.07905934998345]
  Vision-Language-Action (VLA) models revolutionize robotic systems by enabling end-to-end perception-to-action pipelines.<n>These pipelines integrate multiple sensory modalities, such as visual signals processed by cameras and auditory signals captured by microphones.<n>Given the fact that VLA-based systems heavily rely on the sensory input, the security of VLA models against physical-world sensor attacks remains critically underexplored.
  arXiv  Detail & Related papers  (2025-11-13T06:24:28Z)
• DetectiumFire: A Comprehensive Multi-modal Dataset Bridging Vision and Language for Fire Understanding [5.257894673786823]
  We introduce DetectiumFire, a large-scale, multi-modal dataset comprising of 22.5k high-resolution fire-related images and 2.5k real-world fire-related videos.<n>The data are annotated with both traditional computer vision labels (e.g., bounding boxes) and detailed textual prompts describing the scene.<n>We validate the utility of DetectiumFire across multiple tasks, including object detection, diffusion-based image generation, and vision-language reasoning.
  arXiv  Detail & Related papers  (2025-11-04T11:33:11Z)
• Uncertainty-Aware Post-Detection Framework for Enhanced Fire and Smoke Detection in Compact Deep Learning Models [0.0]
  Existing vision-based methods face challenges in balancing efficiency and reliability.<n>Deep learning models such as YOLOv5n and YOLOv8n are widely adopted for deployment on UAVs, CCTV systems, and IoT devices.<n>We propose an uncertainty aware post-detection framework that rescales detection confidences using both statistical uncertainty and domain relevant visual cues.
  arXiv  Detail & Related papers  (2025-10-11T08:36:57Z)
• Universal Camouflage Attack on Vision-Language Models for Autonomous Driving [67.34987318443761]
  Visual language modeling for automated driving is emerging as a promising research direction.<n>VLM-AD remains vulnerable to serious security threats from adversarial attacks.<n>We propose the first Universal Camouflage Attack framework for VLM-AD.
  arXiv  Detail & Related papers  (2025-09-24T14:52:01Z)
• Robust Anti-Backdoor Instruction Tuning in LVLMs [53.766434746801366]
  We introduce a lightweight, certified-agnostic defense framework for large visual language models (LVLMs)<n>Our framework finetunes only adapter modules and text embedding layers under instruction tuning.<n>Experiments against seven attacks on Flickr30k and MSCOCO demonstrate that ours reduces their attack success rate to nearly zero.
  arXiv  Detail & Related papers  (2025-06-04T01:23:35Z)
• Real-Time Aerial Fire Detection on Resource-Constrained Devices Using Knowledge Distillation [29.875425833515973]
  Wildfire catastrophes cause significant environmental degradation, human losses, and financial damage.<n>Current systems rely primarily on fixed CCTV cameras with a limited field of view, restricting their effectiveness in large outdoor environments.<n>We present a lightweight fire detection model based on MobileViT-S, compressed through the distillation of knowledge from a stronger teacher model.
  arXiv  Detail & Related papers  (2025-02-28T11:49:58Z)
• MarvelOVD: Marrying Object Recognition and Vision-Language Models for Robust Open-Vocabulary Object Detection [107.15164718585666]
  We investigate the root cause of VLMs' biased prediction under the open vocabulary detection context. Our observations lead to a simple yet effective paradigm, coded MarvelOVD, that generates significantly better training targets. Our method outperforms the other state-of-the-arts by significant margins.
  arXiv  Detail & Related papers  (2024-07-31T09:23:57Z)
• Meta-UDA: Unsupervised Domain Adaptive Thermal Object Detection using Meta-Learning [64.92447072894055]
  Infrared (IR) cameras are robust under adverse illumination and lighting conditions. We propose an algorithm meta-learning framework to improve existing UDA methods. We produce a state-of-the-art thermal detector for the KAIST and DSIAC datasets.
  arXiv  Detail & Related papers  (2021-10-07T02:28:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.