Dynamic Density Topological Structure Generation for Real-Time Ladder Affordance Detection

Robots typically possess sensors of different modalities, such as colour cameras, inertial measurement units, and 3D laser scanners. Often, solving a particular problem becomes easier when more than one modality is used. However, while there are undeniable benefits to combine sensors of different modalities the process tends to be complicated. Segmenting scenes observed by the robot into a discrete set of classes is a central requirement for autonomy as understanding the scene is the first step to reason about future situations. Scene segmentation is commonly performed using either image data or 3D point cloud data. In computer vision many successful methods for scene segmentation are based on conditional random fields (CRF) where the maximum a posteriori (MAP) solution to the segmentation can be obtained by inference. In this paper we devise a new CRF inference method for scene segmentation that incorporates global constraints, enforcing the sets of nodes are assigned the same class label. To do this efficiently, the CRF is formulated as a relaxed quadratic program whose MAP solution is found using a gradient-based optimisation approach. The proposed method is evaluated on images and 3D point cloud data gathered in urban environments where image data provides the appearance features needed by the CRF, while the 3D point cloud data provides global spatial constraints over sets of nodes. Comparisons with belief propagation, conventional quadratic programming relaxation, and higher order potential CRF show the benefits of the proposed method.

A novel parameterization concept for structural truss topology optimization is presented in this article that enables the use of evolutionary algorithms in design of large-scale structures. The representational power of Boolean networks is used here to parameterize truss topology. A genetic algorithm then operates on parameters that govern the generation of truss topologies using this random network instead of operating directly on design variables. A genetic algorithm implementation is also presented that is congruent with the local rule application of the random network. The primary advantage of using a Boolean random network representation is that a relatively large number of ground structure nodes can be used, enabling successful exploration of a large-scale design space. In the classical binary representation of ground structures, the number of optimization variables increases quadratically with the number of nodes, restricting the maximum number of nodes that can be considered using a ground structure approach. The Boolean random network representation proposed here allows for the exploration of the entire topology space in a systematic way using only a linear number of variables. The number of nodes in the design domain, therefore, can be increased significantly. Truss member geometry and size optimization is performed here in a nested manner where an inner loop size optimization problem is solved for every candidate topology using sequential linear programming with move-limits. The Boolean random network and nested inner-loop optimization allows for the concurrent optimization of truss topology, geometry, and size. The effectiveness of this method is demonstrated using a planar truss design optimization benchmark problem.

Discrimination of singleton and periodic attractors in Boolean networks

This paper discusses a mechanism for integrating locomotion with cognition in robots. We demonstrate an attentional ability model that can dynamically change the focus of its perceptual area by integrating attention and perception to generate behavior. The proposed model considers both internal sensory information and also external sensory information. We also propose affordance detection that identifies different actions depending on the robot’s immediate possibilities. Attention is represented in a topological structure generated by a growing neural gas that uses 3D point-cloud data. When the robot faces an obstacle, the topological map density increases in the suspected obstacle area. From here, affordance information is processed directly into the behavior pattern generator, which comprises interconnections between motor and internal sensory neurons. The attention model increases the density associated with the suspected obstacle to produce a detailed representation of the obstacle. Then, the robot processes the cognitive information to enact a short-term adaptation to its locomotion by changing its swing pattern or movement plan. To test the effectiveness of the proposed model, it is implemented in a computer simulation and also in a medium-sized, four-legged robot. The experiments validate the advantages in three categories: (1) Development of attention model using topological structure, (2) Integration between attention and affordance in moving behavior, (3) Integration of exteroceptive sensory information to lower-level control of locomotion generator.

This paper presents a new feature descriptor for real-time people detection in depth images. The shape cue in depth images can reduce negative impacts of variations of clothing, lighting conditions and the complexity of backgrounds. The proposed Simplified Local Ternary Patterns (SLTP) can take advantage of depth images to describe human body shape with low computational cost. To evaluate the SLTP feature, we establish a dataset with 7260 positive samples. A series of experiments are carried out on this dataset, and the results show that the SLTP feature can achieve a high detection rate with a low false positive rate. Besides, SLTP is easy to implement, and performs fast (over 80 frames per second) on a standard desktop computer.

RPEA: A Residual Path Network with Efficient Attention for 3D pedestrian detection from LiDAR point clouds

We investigate the dynamics of a susceptible infected recovered (SIR) epidemic model on small networks with different topologies, as a stepping stone to determining how the structure of a contact network impacts the transmission of infection through a population. For an SIR model on a network of$N$nodes, there are$3^{N}$configurations that the network can be in. To simplify the analysis, we group the states together based on the number of nodes in each infection state and the symmetries of the network. We derive analytical expressions for the final epidemic size of an SIR model on small networks composed of three or four nodes with different topological structures. Differential equations which describe the transition of the network between states are also derived and solved numerically to confirm our analysis. A stochastic SIR model is numerically simulated on each of the small networks with the same initial conditions and infection parameters to confirm our results independently. We show that the structure of the network, degree of the initial infectious node, number of initial infectious nodes and the transmission rate all significantly impact the final epidemic size of an SIR model on small networks.

Controlling the dynamics of biological neuronal network (BNNs) is of great importance for both understanding the brain and for developing means to control brain disease related aberrant activity. In the last decade, the relationship among node properties, network structure and network dynamics has become clearer and several neuron and network properties could be used to control the network dynamics. Designing appropriate control inputs would require selective manipulation of the control variable in each node. However, the large size of BNNs and scarcity of resources do not allow control of all individual nodes by external means. Therefore, there is a need to identify a much smaller number of influential nodes that are sufficient to drive the activity dynamics of the entire BNN to a desired state. Here, we used the concept of structural controllability [1] to identify a small set of influential nodes, control nodes, for synthetic and experimental networks. The later includes large-scale connectivity of macaque cortical regions [2], meso-scale connectome of a mouse brain [3] and the neuronal connectivity of C. elegans [4]. We investigated the effect of random and systematic edge pruning on the number of these control nodes. Moreover, We showed that there is a power law or exponential dependence between the number of control nodes and the average degree of many networks. Finally, we used a simplified network of the basal ganglia to qualitatively show that few control nodes can indeed drive the network dynamics to a desired state. While every sub-network could be used as a control node or sensor, each node is different in terms of the required stimulation pattern, stimulation energy and time to reach the desired dynamical state.

Smoothing topology optimization results using pre-built lookup tables

We develop a widely applicable algorithm to solve the fault diagnosis problem in certain distributed-memory multiprocessor systems in which there are a limited number of faulty processors. In particular, we prove that if the underlying graph forming the interconnection network has connectivity no less than its diagnosability ¿ and can be partitioned into enough connected components of large enough size then given a syndrome of test results under the comparison diagnosis model resulting from some set of faulty nodes of size at most ¿, we can find the actual set of faulty nodes with time complexity O(¿N), where ¿ is the maximal degree of any node of the graph and N is the number of nodes.

With the advancements in information, communication, and sensing technologies, structural health monitoring (SHM) has matured into a substantial pillar of infrastructure maintenance. In particular, wireless sensor networks have gradually been incorporated into SHM, leveraging new opportunities towards reduced installation efforts and enhanced flexibility and scalability, as compared to cable-based SHM systems. However, wireless sensor nodes are installed at fixed locations and need to be employed at high density to reliably monitor large infrastructure, which may cause high installation costs. Furthermore, the limited power autonomy of wireless sensor networks, installed at fixed locations for unattended long-term operation, still represents a significant constraint when deploying stationary wireless sensor nodes for SHM. To resolve the critical constraints stemming from costly high-density deployment and limited power autonomy, a mobile structural health monitoring concept based on legged robots is proposed in the study reported in this paper. The study explores the accuracy and cost-efficiency of deploying legged robots in dense measurement setups for wireless SHM of civil infrastructure, aiming to gain insights into the advantages of mobile wireless sensor nodes in general and of legged robots in particular, in terms of obtaining rich information on the structural condition. As is shown in this paper, the legged robots, as compared to stationary wireless sensor nodes, require a smaller number of nodes to be deployed in civil infrastructure to achieve rich sensor information, entailing more cost-efficient, yet accurate, SHM. In conclusion, this study represents a first step towards autonomous robotic fleets advancing structural health monitoring.

The philosophy behind the modular design and implementation of a computer-driven facility to collect and process EEG records is illustrated. The software modules are stand-alone programs in their own right that can externally be reconfigured into the components to solve a more complex problem. As such, they behave like the commands of a command line interpreter (CLI). They are easily invoked by the use of operating system macros without increasing the demands on the system operator. By the use of an experiment parameter and specifications table (a common data base) the specifics of a given series of experiments are prestored well in advance of data collection; specification of a new table filename changes the nature of the experiment. Complete operating system macro menus can be predesigned and stored. To change the experiment the operator simply changes the menu. The system has been used to collect and process single visual evoked responses (unaveraged EEG records) elicited from human subjects for a series of cognitive experiments with relatively good classification error rates.

Apple detection plays a critical role in enabling the functionality of harvesting robots within natural orchard environments. To address challenges related to low detection accuracy, slow inference speed, and high parameter count, we present PcMNet, a lightweight detection model based on an improved YOLOv8 network. Initially, we employed Partial Convolution (Pconv) to construct a PR module, forming the Pconv-block, which subsequently replaced the original C2f feature extraction module within the YOLOv8n backbone. This replacement led to improvements in both detection accuracy and speed, while simultaneously reducing computational complexity (FLOPs), parameter count, and model size. Furthermore, the Cross-Scale Feature Fusion (CCFF) module was refined into Faster-Cross-Scale Feature Fusion (Faster-CCFF) with the integration of Pconv-block, significantly enhancing the model's feature extraction and fusion capabilities. Additionally, we introduced Mixed Local Channel Attention (MLCA) to further strengthen the model's capacity to capture essential features while effectively suppressing background noise. Experimental results demonstrate that PcMNet achieved a detection accuracy of 92.8 % and an mAP@0.5 of 95.5 %, representing improvements of 1.4 and 0.7 percentage points, respectively, over YOLOv8n. Moreover, PcMNet successfully reduced FLOPs, parameter count, and model size to 5.1 G, 1.4 M, and 3.2 MB, respectively. The per-image detection time was reduced to 2.3 ms, indicating reductions of 37.80 %, 53.33 %, 49.21 %, and 56.60 % in FLOPs, parameters, model size, and detection time compared to YOLOv8n. When deployed on edge computing devices with TensorRT acceleration, PcMNet achieved a detection rate of 92 FPS. Field validation experiments conducted in natural orchard environments confirmed PcMNet's superior ability to detect apples under challenging conditions, such as occlusions and varying lighting conditions. Its lightweight design and rapid detection capabilities provide a valuable reference for achieving real-time apple detection in automated and intelligent harvesting robots, thereby contributing to advancements in smart agriculture.

Monte Carlo Methods for Node Self-Localization and Nonlinear Target Tracking in Wireless Sensor Networks

The goal of the t/k-diagnosis is to isolate all faulty processors (nodes) in a multiprocessor system to within a set of nodes in which at most k nodes are correct, provided the number of faulty nodes does not exceed t. As compared to the classical precise diagnosis strategy, the t/k-diagnosis strategy can significantly improve the self-diagnosing capability of multiprocessor system. The generalised cube network (GCN), or equivalently the BC graphs, is a regular topology, which provides a unified view of the hypercube and some of its variants. This paper addressed the t/k-diagnosis of GCNs. By exploring the relationship between the size of a largest connected component of the 0-test subgraph of a faulty GCN and the distribution of the faulty nodes over the network, an time (4n − 9)/3 diagnosis algorithm on an n-dimensional GCN is presented, where N = 2 n is the total number of the nodes of the network being diagnosed. To our knowledge, this is the first time to give a t/k-diagnosis algorithm for GCNs and .