Accurate target perception in unstructured outdoor environments remains a fundamental challenge in computational imaging and machine vision, primarily due to severe optical degradation caused by variable illumination, specular highlights, and dense foliage occlusion. Existing optical sensing systems often struggle to maintain robustness under these physical constraints, especially when deployed on edge devices with strict computational limits. To address these challenges, this paper proposes Orchard-YOLO, a lightweight, computationally efficient object detection network designed to maintain robustness against environmental and optical noise in complex orchard environments. Unlike generic architectures, Orchard-YOLO introduces three architectural enhancements for robust detection: (1) a High-Resolution P2 Detection Head to preserve high-frequency optical details and fine-grained texture cues often lost during digital downsampling; (2) Coordinate Attention (CA) mechanisms integrated into the feature fusion pathway to filter out background optical interference and enhance spatial discrimination for heavily occluded targets; and (3) a Ghost-convolution-based backbone to optimize the inference pipeline for real-time edge processing. Evaluated on a comprehensive multi-fruit dataset under simulated optical stress (including ±50% illumination variation and up to 70% occlusion), Orchard-YOLO achieves 94.8% mAP@0.5. It shows improved robustness under illumination variation and occlusion compared to baseline models, while achieving up to 25 FPS on an NVIDIA Jetson Nano edge device. These results suggest that Orchard-YOLO offers a detection framework suitable for resource-constrained orchard perception.
Coral reefs are facing a catastrophic decline due to climate-induced bleaching, threatening critical marine biodiversity. Automated, large-scale monitoring is essential; however, modern object detectors are hindered by two fundamental limitations in complex underwater scenes: a spatial reasoning deficit in their decoupled heads, which inhibits robust multi-scale feature integration, and a feature robustness deficit, which renders deterministic networks vulnerable to stochastic visual variations. To address these limitations, we propose Coral-YOLO, a novel framework for detection and forecasting. We introduce the Holistic Attention Block Head (HAB-Head), which enables deep cross-scale reasoning through explicit feature interaction, and MCAttention, a randomized training mechanism that enables the network to learn scale-invariant features with inherent robustness. Evaluated on our newly curated, multi-year CR-Mix dataset, Coral-YOLO achieves a state-of-the-art 50.3% AP (average precision at IoU threshold 0.5:0.95, following COCO metrics), representing a +1.8 percentage point improvement over the YOLOv12-m baseline, with particularly pronounced gains on small objects (+2.6 percentage points in ). Crucially, its integrated temporal forecasting module achieves 82.7% accuracy in predicting future coral health, substantially outperforming conventional methods. Coral-YOLO sets a new performance benchmark and enables proactive reef conservation. It provides a powerful tool to identify at-risk corals long before severe bleaching becomes visually apparent.
The autonomous coordination of multi-robot systems in complex, environments remains a fundamental challenge. Current Multi-Agent Reinforcement Learning (MARL) methods often struggle to reason effectively about the dynamic, causal relationships between agents and their surroundings. To address this, we introduce the Graph-Gated Transformer (GGT), a novel neural architecture designed to inject explicit relational priors directly into the self-attention mechanism for multi-robot coordination. The core mechanism of the GGT involves dynamically constructing a Tactical Relational Graph that encodes high-priority relationships like collision risk and cooperative intent. This graph is then used to generate an explicit attention mask, compelling the Transformer to focus its reasoning exclusively on entities rather than engaging in brute-force pattern matching across all perceived objects. Integrated into a Centralized Training with Decentralized Execution (CTDE) framework with QMIX, our approach demonstrates substantial improvements in high-fidelity simulations. In complex scenarios with dynamic obstacles and sensor noise, our GGT-based system achieves 95.3% coverage area efficiency with only 0.4 collisions per episode, a stark contrast to the 60.3% coverage and 20.7 collisions of standard QMIX. Ablation studies confirm that this structured, gated attention mechanism—not merely the presence of attention—is the key to unlocking robust collective autonomy. This work establishes that explicitly constraining the Transformer’s attention space with dynamic, domain-aware relational graphs is a powerful and effective architectural solution for engineering safe and intelligent multi-robot systems.
In maritime navigation, vessel dynamics and open-water environments often exhibit inherent symmetries, whereas control objectives, particularly collision avoidance and COLREGs compliance, are strictly asymmetric, specifying unique responsibilities (e.g., give-way versus stand-on) and distinct desired trajectories. This paper proposes a runtime-assured, dual-envelope predictive safety-control framework for autonomous surface vehicles (ASVs) that directly addresses the symmetry and asymmetry in complex encounters. To manage asymmetric semantic objectives, a large language model (LLM) serves as a semantic-governance module, generating structured COLREGs labels (encounter type, vessel responsibility, and maneuver tendency). These semantic outputs are strictly validated before entering the control stack. In parallel, to break the dangerous symmetry of collision risks, vessel-motion prediction and uncertainty inflation construct a physical safety envelope. A deterministic MPC-CBF safety filter then computes admissible control commands, balancing the symmetric homogeneous tracking dynamics with asymmetric collision-avoidance constraints. A runtime assurance monitor supervises semantic validity and solver latency, preventing unsafe decisions. Simulation results demonstrate that the proposed intelligent decision-making and control scheme significantly improves rule-aware collision avoidance while reducing excessive conservatism, providing a profound perspective for viewing maritime safety through symmetric and asymmetric control theory.
Against the backdrop of escalating marine pollution, automated visual detection of submerged debris by autonomous underwater vehicles (AUVs) is essential for robotic ocean cleanup. However, robust underwater detection is fundamentally challenged by an asymmetric visual degradation process: wavelength-dependent light attenuation weakens red channel responses more severely than blue–green channels, turbidity-induced scattering disrupts the symmetry of small target features, and cross-water-domain shifts further distort feature distributions. To address these challenges, this paper proposes AquaYOLO26, a degradation-aware and domain-adaptive edge detection framework that explicitly models underwater asymmetry while restoring feature-level balance for robust perception. Specifically, we introduce Underwater-aware Batch Normalization (UWBN) to compensate for channel-asymmetric color attenuation within the feature extractor; Turbidity-conditioned Small Target-Aware Label Assignment (Turb-STAL) to asymmetrically adjust candidate assignment for obscured small targets under scattering conditions; and Domain-Adversarial Progressive Loss (DA-ProgLoss) to align feature representations across heterogeneous water environments. Experiments on the SeaClear, TrashCan 1.0, and Trash-ICRA19 benchmarks show that AquaYOLO26n achieves 73.4 ± 0.33% mAP@0.5 on the primary SeaClear test split and, in the representative run cross-domain evaluation, improves strict zero-shot transfer performance on Trash-ICRA19 by 8.9 percentage points over the YOLO26n baseline. For embedded deployment, the model retains a compact profile of 2.46 M parameters and 6.2 GFLOPs, achieving 45.1 ± 0.5 FPS on the NVIDIA Jetson Orin NX. Compared with Improved YOLOv11, AquaYOLO26n achieves a 1.1 pp higher mean mAP@0.5 while using 79.0% fewer parameters and 77.0% fewer GFLOPs and delivering higher Jetson throughput (45.1 ± 0.5 vs. 28.4 ± 0.5 FPS). These results highlight a favorable accuracy–efficiency trade-off for resource-constrained underwater robotic sensing.
Dense underwater aquaculture poses significant challenges for intelligent image processing because asymmetric occlusion, turbidity, aeration-like bubbles, and motion blur frequently degrade fish contours and quasi-periodic scale textures. These disturbances often cause conventional detectors to miss detections, merge bounding boxes, experience feature collapse, and exhibit unstable counting. To address this problem, we propose DenseFish-v13, a symmetry-aware NMS-free YOLOv13-Mamba framework for dense underwater fish detection and bio-kinematic behavior recognition. The framework integrates a Bio-Harmonic Frequency Gate to preserve biological texture patterns while suppressing bubble-like frequency noise, a Bi-directional Multi-scale Wavelet Mamba backbone for global occlusion-aware structure recovery, and an asymmetry-aware density repulsion strategy to separate highly overlapping fish instances during bipartite matching. In addition, a lightweight Bio-Kinematic Behavior Head converts continuous detections into interpretable trajectory descriptors for behavior-state recognition. Experiments on the Dense-Aqua benchmark, constructed from public aquaculture datasets, show that DenseFish-v13 achieves 64.8% mAP@50:95 and a Counting MAE of 3.7 on the overall test set, while reaching 64.2% mAP@50:95 and a Counting MAE of 4.1 on the extreme-density split. Under a strong synthetic bubble perturbation, the model shows only a 1.3 percentage-point drop in mAP and maintains 125 FPS on Jetson Orin NX. These results demonstrate its effectiveness in robust, real-time underwater aquaculture monitoring.