Frontiers in Neurorobotics最新文献

[This corrects the article DOI: 10.3389/fnbot.2025.1574044.].

Introduction: The combination of CNN and Transformer has attracted much attention for medical image segmentation due to its superior performance at present. However, the segmentation performance is affected by limitations such as the local receptive field and static weights of CNN convolution operations, as well as insufficient information exchange between Transformer local regions.

Methods: To address these issues, an integrated attention mechanism and pyramid pooling network is proposed in this paper. Firstly, an efficient channel attention mechanism is embedded into CNN to extract more comprehensive image features. Then, CBAM_ASPP module is introduced into the bottleneck layer to obtain multi-scale context information. Finally, in order to address the limitations of traditional convolution, depthwise separable convolution is used to achieve a lightweight network.

Results: The experiments based on the Synapse multi organ segmentation dataset and ACDC dataset showed that the proposed IAP-TransUNet achieved Dice similarity coefficients (DSCs) of 78.85% and 90.46%, respectively. Compared with the state-of-the-art method, for the Synapse multi organ segmentation dataset, the Hausdorff distance was reduced by 2.92%. For the ACDC dataset, the segmentation accuracy of the left ventricle, myocardium, and right ventricle was improved by 0.14%, 1.89%, and 0.23%, respectively.

Discussion: The experimental results demonstrate that the proposed network has improved the effectiveness and shows strong performance on both CT and MRI data, which suggests its potential for generalization across different medical imaging modalities.

Background: Physiotherapy robots offer a feasible and promising solution for achieving safe and efficient treatment. Among these, acupoint recognition is the core component that ensures the precision of physiotherapy robots. Although the research on the acupoint recognition such as hand and ear has been extensive, the accurate location of acupoints on the back of the human body still faces great challenges due to the lack of significant external features.

Methods: This paper designs a two-stage acupoint recognition method, which is achieved through the cooperation of two detection networks. First, a lightweight RTMDet network is used to extract the effective back range from the image, and then the acupoint coordinates are inferred from the extracted back range, reducing the inference consumption caused by invalid information. In addition, the RTMPose network based on the SimCC framework converts the acupoint coordinate regression problem into a classification problem of sub-pixel block subregions on the X and Y axes by performing sub-pixel-level segmentation of images, significantly improving detection speed and accuracy. Meanwhile, the multi-layer feature fusion of CSPNeXt enhances feature extraction capabilities. Then, we designed a physiotherapy interaction interface. Through the three-dimensional coordinates of the acupoints, we independently planned the physiotherapy task path of the physiotherapy robot.

Results: We conducted performance tests on the acupoint recognition system and physiotherapy task planning in the physiotherapy robot system. The experiments have proven our effectiveness, achieving a recall of 90.17% on human datasets, with a detection error of around 5.78 mm. At the same time, it can accurately identify different back postures and achieve an inference speed of 30 FPS on a 4070Ti GPU. Finally, we conducted continuous physiotherapy tasks on multiple acupoints for the user.

Conclusion: The experimental results demonstrate the significant advantages and broad application potential of this method in improving the accuracy and reliability of autonomous acupoint recognition by physiotherapy robots.

C. elegans is a model organism in many biological domains, such as genetics, neurophysiology, and behavioral ecology. Despite our relatively deep knowledge of the neuronal, genetic and molecular mechanisms underlying C. elegans communication, we still lack a comprehensive understanding of emergent group-level dynamics. We review the literature on collective behavior of C. elegans by categorizing works in this relatively small research field along three main axes corresponding to primary collective responses: aggregation, swarming, and collective decision-making. Through an analysis of the methods and scientific contributions of these works, we develop a critical perspective that points to important gaps in our understanding of the mechanisms underlaying the emergence of collective responses. We discuss the consequences of the lack of evidence concerning the effect of population density on the emergence of specific group dynamics, and the relatively limited knowledge related to how self-generated pheromones regulate local interactions and contribute to the emergence of group responses. We elaborate on the methodological problems of developing experimental scenarios to disentangle causal relationships between population density, pheromone-based interactions and collective responses. We propose to overcome these limitations with an interdisciplinary approach based on the use of in vivo experiments, mathematical and computer-based models.

For the low efficiency and poor generalization ability of path planning algorithm of industrial robots, this work proposes an adaptive field co-sampling algorithm (AFCS). Firstly, the environment complexity function is proposed to make full use of environment information and improve its generalization ability of the traditional rapidly random search tree algorithm (RRT) algorithm. Then an optimal sampling strategy is proposed to make the improvement of the efficiency and optimal direction of RRT algorithm. Finally, this article designs a collaborative extension strategy, which introduces the improved artificial potential field algorithm (APF) into the traditional RRT algorithm to determine the new nodes, so as to improve the orientation and expansion efficiency of the algorithm. The proposed AFCS algorithm completes simulation experiments in two environments with different complexity. Compared with the traditional RRT, RRT* and tRRT algorithm, the results show that the AFCS algorithm has achieved great improvement in environmental adaptability, stability and efficiency. At last, ROKAE industrial robot is taken as the object to build a simulation environment for the path planning, which further verifies the practicability of the algorithm.

[This corrects the article DOI: 10.3389/fnbot.2025.1633697.].

Under the influence of Masked Language Modeling (MLM), Masked Image Modeling (MIM) employs an attention mechanism to perform masked training on images. However, processing a single image requires numerous iterations and substantial computational resources to reconstruct the masked regions, resulting in high computational complexity and significant time costs. To address this issue, we propose an Effective and Efficient self-supervised Masked model based on Mixed feature training (EESMM). First, we stack two images for encoding and input the fused features into the network, which not only reduces computational complexity but also enables the learning of more features. Second, during decoding, we obtain the decoding features corresponding to the original images based on the decoding features of the two input original images and the mixed images, and then construct a corresponding loss function to enhance feature representation. EESMM significantly reduces pre-training time without sacrificing accuracy, achieving 83% accuracy on ImageNet in just 363 h using four V100 GPUs-only one-tenth of the training time required by SimMIM. This validates that the method can substantially accelerate the pre-training process without noticeable performance degradation.

During turning maneuvers in the galloping gait of quadruped animals, a strong relationship exists between the turning direction and the sequence in which the forelimbs make ground contact: the outer forelimb acts as the "trailing limb" while the inner forelimb serves as the "leading limb." However, the control mechanisms underlying this behavior remain largely unclear. Understanding these mechanisms could deepen biological knowledge and assist in developing more agile robots. To address this issue, we hypothesized that decentralized interlimb coordination mechanism and trunk movement are essential for the emergence of an inside leading limb in a galloping turn. To test the hypothesis, we developed a quasi-quadruped robot with simplified wheeled hind limbs and variable trunk roll and yaw angles. For forelimb coordination, we implemented a simple decentralized control based on local load-dependent sensory feedback, utilizing trunk roll inclination and yaw bending as turning methods. Our experimental results confirmed that in addition to the decentralized control from previous studies which reproduces animal locomotion in a straight line, adjusting the trunk roll angle spontaneously generates a ground contact sequence similar to gallop turning in quadruped animals. Furthermore, roll inclination showed a greater influence than yaw bending on differentiating the leading and trailing limbs. This study suggests that physical interactions serve as a universal mechanism of locomotor control in both forward and turning movements of quadrupedal animals.

Accurate pedestrian tracking and behavior recognition are essential for intelligent surveillance, smart transportation, and human-computer interaction systems. This paper introduces TSLNet, a Hierarchical Multi-Head Attention-Enabled Two-Stream LSTM Network, designed to overcome challenges such as environmental variability, high-density crowds, and diverse pedestrian movements in real-world video data. TSLNet combines a Two-Stream Convolutional Neural Network (Two-Stream CNN) with Long Short-Term Memory (LSTM) networks to effectively capture spatial and temporal features. The addition of a Multi-Head Attention mechanism allows the model to focus on relevant features in complex environments, while Hierarchical Classifiers within a Multi-Task Learning framework enable the simultaneous recognition of basic and complex behaviors. Experimental results on multiple public and proprietary datasets demonstrate that TSLNet significantly outperforms existing baseline models, achieving higher Accuracy, Precision, Recall, F1-Score, and Mean Average Precision (mAP) in behavior recognition, as well as superior Multiple Object Tracking Accuracy (MOTA) and ID F1 Score (IDF1) in pedestrian tracking. These improvements highlight TSLNet's effectiveness in enhancing tracking and recognition performance.

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.