{"slug": "real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain", "title": "Real-Time Diffusion Relighting as a Neural Sensor: Closing the Low-Light Domain Gap for Robotic Perception", "summary": "A new arXiv paper (2611.17438v1) reports a real-time diffusion relighting system that sets a state-of-the-art 87.1% mAP on the ExDark low-light benchmark and achieves a 100% lap completion rate in the CARLA closed-loop autonomous driving simulator. The authors embed a generative neural sensor in the robot control loop, treating illumination as a generative prior, and run a large-scale image-to-image diffusion model in a background thread that activates only when a drop in detection confidence signals a perceptual uncertainty event. The approach decouples high-latency semantic recovery from low-latency robotic control, addressing the failure of daylight-trained object detectors under dark and adverse weather conditions.", "body_md": "arXiv:2611.17438v1 Announce Type: new \nAbstract: Low-light conditions severely degrade robotic perception: object detectors trained on daylight imagery are prone to failure under dark and adverse weather conditions that fall outside of their training distribution. Furthermore, from an efficiency standpoint, existing Low-Light Image Enhancement (LLIE) methods are designed for offline processing and are therefore impractical to integrate into live control loops. To address these issues, we propose to embed a Generative Neural Sensor within the control loop that treats illumination as a generative prior rather than a signal processing problem, enabling task-relevant semantic recovery. We further propose an Real-Time Perceptual Architecture that decouples high-latency semantic recovery from low-latency robotic control. Rather than blocking the control loop, our system runs a large-scale image-to-image diffusion model in a background thread, and is activated only when a drop in detection confidence signals a perceptual uncertainty event, significantly reducing computational overhead. Our method sets a new state-of-the-art performance on the ExDark low-light benchmark with 87.1% mAP, and under adverse weather conditions, significantly boosts detection rates over competing methods. On a navigation task in a closed-loop autonomous driving simulator (CARLA), we achieve a 100% lap completion rate, establishing a simulation-based proof-of-concept for integrating high-latency generative models with standard robotic control autonomy stacks.", "url": "https://wpnews.pro/news/real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain", "canonical_source": "https://arxiv.org/abs/2611.17438", "published_at": "2026-11-16 05:00:00+00:00", "updated_at": "2026-11-16 05:21:02.372623+00:00", "lang": "en", "topics": ["computer-vision", "autonomous-vehicles", "robotics", "generative-ai", "ai-research"], "entities": ["ExDark", "CARLA", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain", "markdown": "https://wpnews.pro/news/real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain.md", "text": "https://wpnews.pro/news/real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain.txt", "jsonld": "https://wpnews.pro/news/real-time-diffusion-relighting-as-a-neural-sensor-closing-the-low-light-domain.jsonld"}}