Frontiers in Neurorobotics最新文献

Load imbalance is a major performance bottleneck in training mixture-of-experts (MoE) models, as unbalanced expert loads can lead to routing collapse. Most existing approaches address this issue by introducing auxiliary loss functions to balance the load; however, the hyperparameters within these loss functions often need to be tuned for different tasks. Furthermore, increasing the number of activated experts tends to exacerbate load imbalance, while fixing the activation count can reduce the model's confidence in handling difficult tasks. To address these challenges, this paper proposes a dynamically balanced routing strategy that employs a threshold-based dynamic routing algorithm. After each routing step, the method adjusts expert weights to influence the load distribution in the subsequent routing. Unlike loss-function-based balancing methods, our approach operates directly at the routing level, avoiding gradient perturbations that could degrade model quality, while dynamically routing to make more efficient use of computational resources. Experiments on Natural Language Understanding (NLU) benchmarks demonstrate that the proposed method achieves accuracy comparable to top-2 routing, while significantly reducing the load standard deviation (e.g., from 12.25 to 1.18 on MNLI). In addition, threshold-based dynamic expert activation reduces model parameters and provides a new perspective for mitigating load imbalance among experts.

Introduction: Traditional methods for farmland road extraction, such as U-Net, often struggle with complex noise and geometric features, leading to discontinuous extraction and insufficient sensitivity. To address these limitations, this study proposes a novel dual-phase generative adversarial network (GAN) named UHGAN, which integrates Hough-transform constraints.

Methods: We designed a cascaded U-Net generator within a two-stage GAN framework. The Stage 1 GAN combines a differentiable Hough transform loss with cross-entropy loss to generate initial road masks. Subsequently, the Stage 2 U-Net refines these masks by repairing breakpoints and suppressing isolated noise.

Results: When evaluated on the WHU RuR+rural road dataset, the proposed UHGAN method achieved an accuracy of 0.826, a recall of 0.750, and an F1-score of 0.789. This represents a significant improvement over the single-stage U-Net (F1 = 0.756) and ResNet (F1 = 0.762) baselines.

Discussion: The results demonstrate that our approach effectively mitigates the issues of discontinuous extraction caused by the complex geometric shapes and partial occlusion characteristic of farmland roads. The integration of Hough-transform loss, an technique that has received limited attention in prior studies, proves to be highly beneficial. This method shows considerable promise for practical applications in rural infrastructure planning and precision agriculture.

Introduction: Urban traffic congestion, environmental degradation, and road safety challenges necessitate intelligent aerial robotic systems capable of real-time adaptive decision-making. Unmanned Aerial Vehicles (UAVs), with their flexible deployment and high vantage point, offer a promising solution for large-scale traffic surveillance in complex urban environments. This study introduces a UAV-based neural framework that addresses challenges such as asymmetric vehicle motion, scale variations, and spatial inconsistencies in aerial imagery.

Methods: The proposed system integrates a multi-stage pipeline encompassing contrast enhancement and region-based clustering to optimize segmentation while maintaining computational efficiency for resource-constrained UAV platforms. Vehicle detection is carried out using a Recurrent Neural Network (RNN), optimized via a hybrid loss function combining cross-entropy and mean squared error to improve localization and confidence estimation. Upon detection, the system branches into two neural submodules: (i) a classification stream utilizing SURF and BRISK descriptors integrated with a Swin Transformer backbone for precise vehicle categorization, and (ii) a multi-object tracking stream employing DeepSORT, which fuses motion and appearance features within an affinity matrix for robust trajectory association.

Results: Comprehensive evaluation on three benchmark UAV datasets-AU-AIR, UAVDT, and VAID shows consistent and high performance. The model achieved detection precisions of 0.913, 0.930, and 0.920; tracking precisions of 0.901, 0.881, and 0.890; and classification accuracies of 92.14, 92.75, and 91.25%, respectively.

Discussion: These findings highlight the adaptability, robustness, and real-time viability of the proposed architecture in aerial traffic surveillance applications. By effectively integrating detection, classification, and tracking within a unified neural framework, the system contributes significant advancements to intelligent UAV-based traffic monitoring and supports future developments in smart city mobility and decision-making systems.

To enhance the obstacle avoidance performance and autonomous decision-making capabilities of robots in complex dynamic environments, this paper proposes an end-to-end intelligent obstacle avoidance method that integrates deep reinforcement learning, spatiotemporal attention mechanisms, and a Transformer-based architecture. Current mainstream robot obstacle avoidance methods often rely on system architectures with separated perception and decision-making modules, which suffer from issues such as fragmented feature transmission, insufficient environmental modeling, and weak policy generalization. To address these problems, this paper adopts Deep Q-Network (DQN) as the core of reinforcement learning, guiding the robot to autonomously learn optimal obstacle avoidance strategies through interaction with the environment, effectively handling continuous decision-making problems in dynamic and uncertain scenarios. To overcome the limitations of traditional perception mechanisms in modeling the temporal evolution of obstacles, a spatiotemporal attention mechanism is introduced, jointly modeling spatial positional relationships and historical motion trajectories to enhance the model's perception of critical obstacle areas and potential collision risks. Furthermore, an end-to-end Transformer-based perception-decision architecture is designed, utilizing multi-head self-attention to perform high-dimensional feature modeling on multi-modal input information (such as LiDAR and depth images), and generating action policies through a decoding module. This completely eliminates the need for manual feature engineering and intermediate state modeling, constructing an integrated learning process of perception and decision-making. Experiments conducted in several typical obstacle avoidance simulation environments demonstrate that the proposed method outperforms existing mainstream deep reinforcement learning approaches in terms of obstacle avoidance success rate, path optimization, and policy convergence speed. It exhibits good stability and generalization capabilities, showing broad application prospects for deployment in real-world complex environments.

Overcoming visual degradation in challenging imaging scenarios is essential for accurate scene understanding. Although deep learning methods have integrated various perceptual capabilities and achieved remarkable progress, their high computational cost limits practical deployment under resource-constrained conditions. Moreover, when confronted with diverse degradation types, existing methods often fail to effectively model the inconsistent attenuation across color channels and spatial regions. To tackle these challenges, we propose DWMamba, a degradation-aware and weight-efficient Mamba network for image quality enhancement. Specifically, DWMamba introduces an Adaptive State Space Module (ASSM) that employs a dual-stream channel monitoring mechanism and a soft fusion strategy to capture global dependencies. With linear computational complexity, ASSM strengthens the models ability to address non-uniform degradations. In addition, by leveraging explicit edge priors and region partitioning as guidance, we design a Structure-guided Residual Fusion (SGRF) module to selectively fuse shallow and deep features, thereby restoring degraded details and enhancing low-light textures. Extensive experiments demonstrate that the proposed network delivers superior qualitative and quantitative performance, with strong generalization to diverse extreme lighting conditions. The code is available at https://github.com/WindySprint/DWMamba.

Introduction: Abandonment rates for myoelectric upper limb prostheses can reach 44%, negatively affecting quality of life and increasing the risk of injury due to compensatory movements. Traditional myoelectric prostheses rely on conventional signal processing for the detection and classification of movement intentions, whereas machine learning offers more robust and complex control through pattern recognition. However, the non-stationary nature of surface electromyogram signals and their day-to-day variations significantly degrade the classification performance of machine learning algorithms. Although single-session classification accuracies exceeding 99% have been reported for 8-class datasets, multisession accuracies typically decrease by 23% between morning and afternoon sessions. Retraining or adaptation can mitigate this accuracy loss.

Methods: This study evaluates three paradigms for retraining a machine learning-based classifier: confidence scores, nearest neighbour window assessment, and a novel signal-to-noise ratio-based approach.

Results: The results show that all paradigms improve accuracy against no retraining, with the nearest neighbour and signal-to-noise ratio methods showing an average improvement 5% in accuracy over the confidence-based approach.

Discussion: The effectiveness of each paradigm is assessed based on intersession accuracy across 10 sessions recorded over 5 days using the NinaPro 6 dataset.

[This corrects the article DOI: 10.3389/fnbot.2025.1604453.].

Introduction: To address the challenges of cumulative errors, insufficient modeling of complex spatiotemporal features, and limitations in computational efficiency and generalization ability in 4D trajectory prediction, this paper proposes a high-precision, robust prediction method.

Methods: A hybrid model SVMD-DBO-RCBAM is constructed, integrating sequential variational modal decomposition (SVMD), the dung beetle optimization algorithm (DBO), and the ResNet-CBAM network. Innovations include frequency-domain feature decoupling, dynamic parameter optimization, and enhanced spatio-temporal feature focusing.

Results: Experiments show that the model achieves a low longitude MAE of 0.0377 in single-step prediction, a 38.5% reduction compared to the baseline model; in multi-step prediction, the longitude R2 reaches 0.9844, with a 72.9% reduction in cumulative error rate and an IQR of prediction errors less than 10% of traditional models, demonstrating high accuracy and stability.

Discussion: Experiments show that the model achieves a low longitude MAE of 0.0377 in single-step prediction, a 38.5% reduction compared to the baseline model; in multi-step prediction, the longitude R2 reaches 0.9844, with a 72.9% reduction in cumulative error rate and an IQR of prediction errors less than 10% of traditional models, demonstrating high accuracy and stability.

With the advancement of deep learning, road crack segmentation has become increasingly crucial for intelligent transportation safety. Despite notable progress, existing methods still face challenges in capturing fine-grained textures in small crack regions, handling blurred edges and significant width variations, and performing multi-class segmentation. Moreover, the high computational cost of training such models hinders their practical deployment. To tackle these limitations, we propose RSA-TransUNet, a novel model for road crack segmentation. At its core is the Axial-shift MLP Attention (ASMA) mechanism, which integrates axial perception with sparse contextual modeling. Through multi-path axial perturbations and an attention-guided structure, ASMA effectively captures long-range dependencies within row-column patterns, enabling detailed modeling of multi-scale crack features. To improve the model's adaptability to structural irregularities, we introduce the Adaptive Spline Linear Unit (ASLU), which enhances the model's capacity to represent nonlinear transformations. ASLU improves responsiveness to microstructural variations, morphological distortions, and local discontinuities, thereby boosting robustness across different domains. We further develop a Structure-aware Multi-stage Evolutionary Optimization (SMEO) strategy, which guides the training process through three phases: structural perception exploration, feature stability enhancement, and global perturbation. This strategy combines breadth sampling, convergence compression, and local escape mechanisms to improve convergence speed, global search efficiency, and generalization performance. Extensive evaluations on the Crack500, CFD, and DeepCrack datasets-including ablation studies and comparative experiments-demonstrate that RSA-TransUNet achieves superior segmentation accuracy and robustness in complex road environments, highlighting its potential for real-world applications.

Although deep learning methods have made significant strides in single image sand dust removal, the heterogeneous uncertainty induced by dusty environments poses a considerable challenge. In response, our research presents a novel framework known as the Hierarchical Interactive Uncertainty-aware Network (HIUNet). HIUNet leverages Bayesian neural networks for the extraction of robust shallow features, bolstered by pre-trained encoders for feature extraction and the agility of lightweight decoders for preliminary image reconstitution. Subsequently, a feature frequency selection mechanism is activated to enhance overall performance by strategically identifying and retaining valuable features while effectively suppressing redundant and irrelevant ones. Following this, a feature enhancement module is applied to the preliminary restoration. This intricate fusion culminates in the production of a restored image of superior quality. Our extensive experiments, using our proposed Sand11K dataset that exhibits various levels of degradation from dust and sand, confirm the effectiveness and soundness of our proposed method. By modeling uncertainty via Bayesian neural networks to extract robust shallow features and selecting valuable features through frequency selection, HIUNet can reconstruct high-quality clean images. For future work, we plan to extend our uncertainty-aware framework to handle extreme sand scenarios.