Sep 2026· IEEE Internet of Things Journal· Vol 13, pp. 39626-39640· 0 citations· 65 references
Abstract
In Internet of Things (IoT) systems, such as autonomous vehicles and robots, visual perception under low-light conditions is often severely degraded due to insufficient illumination and intensive sensor noise. While numerous methods have been developed for low-light enhancement, their reliance on paired annotation and computationally heavy architectures limits their applicability for downstream perception tasks, especially on resource-constrained platforms. To address these issues, this article proposes REZS-Edge, a novel zero-shot enhancement framework that eliminates the need for paired supervision and is specifically designed for efficient edge execution. First, by integrating Retinex theory with the noise-to-noise paradigm, the method jointly estimates illumination conditions and suppresses noise through a unified pipeline. Then, a zero-shot training strategy with global and local illumination constraints facilitates fully unsupervised optimization and enhances generalization. Furthermore, the framework incorporates structural re-parameterization to enable model acceleration at the inference phase while maintaining performance. Extensive experiments demonstrate that REZS-Edge achieves state-of-the-art results on low-light image enhancement (LLIE) datasets, including LOLv1 and LOLv2-Real, with a notably fast inference speed of 6.25 ms on a commercial SoC platform Snapdragon 8 Gen 3. More importantly, the method exhibits strong generalization on critical downstream tasks such as low-light object detection, instance segmentation, and pose estimation. These advantages make REZS-Edge particularly valuable for autonomous driving and robotic systems that require reliable, efficient, and all-weather visual perception, offering a practical and deployable solution for real-time enhancement under challenging lighting conditions.
Visual sensors are fundamental components in intelligent surveillance and autonomous transportation systems. However, in low-illumination scenarios, the reliability of these sensors is severely compromised by physical limitations, such as signal attenuation, high noise levels, and reduced dynamic range, which degrade the performance of downstream perception tasks. To address these challenges, this paper proposes a robust visual sensing framework that integrates adaptive signal enhancement with efficient object detection. object detection. First, to mitigate sensor data degradation, we introduce an improved Zero-Reference Deep Curve Estimation (Zero-DCE) algorithm. This module incorporates anisotropic diffusion for noise suppression while preserving structural edges and integrates a non-local attention mechanism to capture long-range dependencies in the sensor inputs. A dynamic exposure loss function is further employed to adaptively adjust brightness. Second, to ensure efficient interpretation of the enhanced data, we optimize the YOLOv8 architecture by introducing a lightweight C2-star block and the MPDIoU loss function, balancing detection accuracy with the efficient requirements of edge sensing devices. Experimental results on benchmark datasets demonstrate that the proposed framework significantly improves signal quality (PSNR of 17.44 dB on LIME) and detection precision (improving mAP by 1.8%). This work provides an effective solution for robust pedestrian monitoring in adverse lighting conditions, enhancing the environmental perception capabilities of intelligent sensing systems.
Dingnan Zhang, Enqi Yue, Boyang Liu et al.· IEEE Access· 0 citations
Extremely low-light RAW enhancement aims to recover severely attenuated sensor signals, yet existing methods often focus on illumination and noise while overlooking the motion-induced degradations inherent in practical low-light imaging. We present a framework for robust extremely low-light RAW enhancement under realistic acquisition degradations. First, we introduce See in the Degraded Extremely Dark (SIDED), a new dataset that applies controlled motion degradation to extremely low-light RAW pairs while retaining their original sensor noise. Second, we propose a unified RAW tokenizer equipped with explicit domain-conditioned representation calibration to align extremely low-light and well-exposed RAW data, followed by a MeanFlow that performs enhancement in a single function evaluation. To our knowledge, this is the first work to formulate extremely low-light RAW enhancement under realistic motion-degraded acquisition and address it with MeanFlow. We further introduce a physics-guided refinement model to strengthen illumination--reflectance consistency, pixel fidelity, and color preservation without incurring additional inference cost. Extensive experiments demonstrate that our framework achieves state-of-the-art performance in extremely low-light RAW enhancement, and robustly handles coupled motion and noise degradations.
Zepu Wang, Jin Liang, Weijie Xiao et al.· 0 citations
Comprehensive comparisons with CNN-, Transformer- and prevailing Mamba-based approaches verify the advantages of SCI-Mamba in visual authenticity, color fidelity and inference speed, and provides a practical low-light enhancement solution for close-proximity non-cooperative space operations.
Yiyong Sun, W. Shan, Shijun Wei et al.· 0 citations
To address the performance degradation of UAV object detection under low-light conditions, we develop an end-to-end object detection network. This proposed method integrates contrastive learning into the detection framework and establishes feature consistency constraints between low-light and normal-light images through a hierarchical contrastive selection encoder. Since the encoder is required only during training and removed during inference, the proposed framework improves feature robustness without introducing additional inference cost. To further improve object detection accuracy, Frequency Guided Dynamic Attention (FGDA) is introduced into the object detection network, focusing on resolving the issue of redundant interference during feature transmission and enhancing feature representation capability. To improve multi-level spatial feature fusion, an Adaptive Gated Dual-Spatial Fusion (AGDSF) module is further developed, which adaptively strengthens target-relevant responses while weakening background noise. According to the experiments on the VisDrone (dark) dataset and a self-collected nighttime UAV-dark dataset illustrate that the proposed method ensures heightened detection accuracy with low computational overhead, complying with the real-time and robustness requirements of UAV perception in low-light contexts.
You Wang, Jiayi Xu, Mengting Lin et al.· Drones· 0 citations
Vision-based perception systems are central to monitoring, inspection, and safety assurance in smart ports. However, adverse weather, particularly rain and fog, degrades image quality, suppresses structural details, and impairs downstream visual perception. Meanwhile, practical port vision systems are often deployed on resource-constrained edge platforms, imposing strict requirements on computational efficiency. To address these challenges, we propose PortVE, a lightweight visual enhancement framework for adverse port weather conditions. PortVE is built on an encoder-decoder architecture and combines degradation-aware Multi-Scale Pooling with Pool-Conv Downsampling and Frequency Directional Modulation. This design captures coarse-to-fine degradation cues, directional structures, and frequency-domain information while maintaining a compact computational profile. Experiments on public benchmarks and a self-collected port-scene synthetic weather dataset demonstrate that PortVE achieves strong restoration performance and efficient inference. It achieves 37.64 dB PSNR on SOTS outdoor and 34.21 dB PSNR on Test2800 with 6.17 M parameters and 42.83 GFLOPs. Downstream object detection experiments further demonstrate that PortVE improves detection robustness under adverse weather. The source code is publicly available at https://github.com/jssc-ZhaoBG/PortVE