Search

Due to labor shortages in specialty crop industries, a need for robotic automation to increase agricultural efficiency and productivity has arisen. Previous manipulation systems perform well in harvesting in uncluttered and structured environments. High tunnel environments are more compact and cluttered in nature, requiring a rethinking of the large form factor systems and grippers. We propose a novel codesigned framework incorporating a global detection camera and a local eye-in-hand camera that demonstrates precise localization of small fruits via closed-loop visual feedback and reliable error handling. Field experiments in high tunnels show our system can reach an average of 85.0\% of cherry tomato fruit in 10.98s on average.

Semantic maps are fundamental for robotics tasks such as navigation and manipulation. They also enable yield prediction and phenotyping in agricultural settings. In this paper, we introduce an efficient and scalable approach for active semantic mapping in horticultural environments, employing a mobile robot manipulator equipped with an RGB-D camera. Our method leverages probabilistic semantic maps to detect semantic targets, generate candidate viewpoints, and compute corresponding information gain. We present an efficient ray-casting strategy and a novel information utility function that accounts for both semantics and occlusions. The proposed approach reduces total runtime by 8% compared to previous baselines. Furthermore, our information metric surpasses other metrics in reducing multi-class entropy and improving surface coverage, particularly in the presence of segmentation noise. Real-world experiments validate our method's effectiveness but also reveal challenges such as depth sensor noise and varying environmental conditions, requiring further research., Comment: Preprint (under review)

We introduce a novel unsteady shear protocol, which we name Rotary Shear (RS), where the flow and vorticity directions are continuously rotated around the velocity gradient direction by imposing two out-of-phase oscillatory shear (OS) in orthogonal directions. We perform numerical simulations of dense suspensions of rigid non-Brownian spherical particles at volume fractions ($\phi$) between 0.40 and 0.55 subject to this new RS protocol and compare to the classical OS protocol. We find that the suspension viscosity displays a similar non-monotonic response as the strain amplitude ($\gamma_0$) is increased: a minimum viscosity is found at an intermediate, volume-fraction dependent strain amplitude. However, the suspension dynamics is different in the new protocol. Unlike the OS protocol, suspensions under RS do not show self-adsorbing states at any $\gamma_0$ and do not undergo the reversible-irreversible transition: the stroboscropic particle dynamics are always diffusive, which we attribute to the fact that the RS protocol is irreversible. To validate this hypothesis, we introduce a reversible-RS (RRS) protocol, a combination of RS and OS, where we rotate the shear direction (as in RS) until it is instantaneously reversed (as in OS), and find the resulting rheology and dynamics to be closer to OS. Detailed microstructure analysis shows that both the OS and RRS protocols result in a contact-free, isotropic to an in-contact, anisotropic microstructure at the dynamically reversible-to-irreversible transition. The RS protocol does not render such a transition, and the dynamics remain diffusive with an in-contact, anisotropic microstructure for all strain amplitudes.

The successful deployment of deep learning-based acoustic echo and noise reduction (AENR) methods in consumer devices has spurred interest in developing low-complexity solutions, while emphasizing the need for robust performance in real-life applications. In this work, we propose a hybrid approach to enhance the state-of-the-art (SOTA) ULCNet model by integrating time alignment and parallel encoder blocks for the model inputs, resulting in better echo reduction and comparable noise reduction performance to existing SOTA methods. We also propose a channel-wise sampling-based feature reorientation method, ensuring robust performance across many challenging scenarios, while maintaining overall low computational and memory requirements., Comment: 5 pages, 4 figures

Deep learning-based methods that jointly perform the task of acoustic echo and noise reduction (AENR) often require high memory and computational resources, making them unsuitable for real-time deployment on low-resource platforms such as embedded devices. We propose a low-complexity hybrid approach for joint AENR by employing a single model to suppress both residual echo and noise components. Specifically, we integrate the state-of-the-art (SOTA) ULCNet model, which was originally proposed to achieve ultra-low complexity noise suppression, in a hybrid system and train it for joint AENR. We show that the proposed approach achieves better echo reduction and comparable noise reduction performance with much lower computational complexity and memory requirements than all considered SOTA methods, at the cost of slight degradation in speech quality., Comment: 5 pages, 2 figures

This paper addresses the challenge of achieving stable adaptive teleoperation and improving the convergence rate in the presence of high communication time delays. We employ a passivity-based formalism to establish stability using wave variables and wave scattering techniques, and we enhance the convergence rate by combining it with predictor-based approaches. The elevated time delay within the teleoperated communication layer is known to induce an oscillatory behavior, which reduces the convergence rate and increases the settling time in the convergence of power variables. This issue is addressed in this paper by utilizing a Smith predictor on the operator end and Minimum Jerk (MJ) predictor on the remote end. We present experimental and simulation results to demonstrate the improvements, ensuring stable teleoperation under high communication time delays., Comment: accepted for publication and presented at ACC 2024

Enhancing the routing efficacy of Flying AdHoc Networks (FANETs), a network of numerous Unmanned Aerial Vehicles (UAVs), in which various challenges may arise as a result of the varied mobility, speed, direction, and rapid topology changes. Given the special features of UAVs, in particular their fast mobility, frequent topology changes, and 3D space movements, it is difficult to transport them through a FANET. The suggested study presents a complete hybrid model: HIROL (Hybrid Intelligent Routing with Optimized Learning) that integrates the ABC (Artificial Bee Colony) algorithm, DSR (Dynamic Source Routing) by incorporating Optimized Link State Routing (OLSR) and ANNs (Artificial Neural Networks) to optimize the routing process. The HIROL optimizes link management by ABC optimization algorithm and reliably analyses link status using characteristics from OLSR and DSR; at the same time, an ANN-based technique successfully classifies connection state. In order to provide optimal route design and maintenance, HIROL dynamically migrates between OLSR and DSR approaches according to the network topology conditions. After running thorough tests in Network Simulator 2 (NS-2), when compared to more conventional DSR and OLSR models, the hybrid model HIROL performs far better in simulations and tests. An increase in throughput (3.5 Mbps vs. 3.2-3.4 Mbps), a decrease in communication overhead (15% vs. 18-20%), and an improvement in Packet Delivery Ratio (97.5% vs. 94-95.5%). These results demonstrate that the suggested HIROL model improves FANET routing performance in different types of networks.

Video foreground segmentation (VFS) is an important computer vision task wherein one aims to segment the objects under motion from the background. Most of the current methods are image-based, i.e., rely only on spatial cues while ignoring motion cues. Therefore, they tend to overfit the training data and don't generalize well to out-of-domain (OOD) distribution. To solve the above problem, prior works exploited several cues such as optical flow, background subtraction mask, etc. However, having a video data with annotations like optical flow is a challenging task. In this paper, we utilize the temporal information and the spatial cues from the video data to improve OOD performance. However, the challenge lies in how we model the temporal information given the video data in an interpretable way creates a very noticeable difference. We therefore devise a strategy that integrates the temporal context of the video in the development of VFS. Our approach give rise to deep learning architectures, namely MUSTAN1 and MUSTAN2 and they are based on the idea of multi-scale temporal context as an attention, i.e., aids our models to learn better representations that are beneficial for VFS. Further, we introduce a new video dataset, namely Indoor Surveillance Dataset (ISD) for VFS. It has multiple annotations on a frame level such as foreground binary mask, depth map, and instance semantic annotations. Therefore, ISD can benefit other computer vision tasks. We validate the efficacy of our architectures and compare the performance with baselines. We demonstrate that proposed methods significantly outperform the benchmark methods on OOD. In addition, the performance of MUSTAN2 is significantly improved on certain video categories on OOD data due to ISD., Comment: 10 pages, 8 figures

The magnetic sublayers introduced by intercalation into the host transition-metal dichalcogenide (TMD) are known to produce various magnetic states. The magnetic sublayers and their magnetic ordering strongly modify the electronic coupling between layers of the host compound. Understanding the roots of this variability is a significant challenge. Here we employ the angle-resolved photoelectron spectroscopy at various photon energies, the {\it ab initio} electronic structure calculations, and modeling to address the particular case of Ni-intercalate, Ni$_{1/3}$NbS$_2$. We find that the bands around the Fermi level bear the signature of a strong yet unusual hybridization between NbS$_2$ conduction band states and the Ni 3$d$ orbitals. The hybridization between metallic NbS$_2$ layers is almost entirely suppressed in the central part of the Brillouin zone, including the part of the Fermi surface around the $\mathrm{\Gamma}$ point. Simultaneously, it gets very pronounced towards the zone edges. It is shown that this behavior is the consequence of the rather exceptional, {\it symmetry imposed}, spatially strongly varying, {\it zero total} hybridization between relevant Ni magnetic orbitals and the neighboring Nb orbitals that constitute the metallic bands. We also report the presence of the so-called $\beta$-feature, discovered only recently in two other magnetic intercalates with very different magnetic orderings. In Ni$_{1/3}$NbS$_2$, the feature shows only at particular photon energies, indicating its bulk origin. Common to prior observations, it appears as a series of very shallow electron pockets at the Fermi level, positioned along the edge of the Brillouin zone. Unforeseen by {\it ab initio} electronic calculations, and its origin still unresolved, the feature appears to be a robust consequence of the intercalation of 2H-NbS$_2$ with magnetic ions., Comment: 14 pages, 7 figures with subfigures, 53 references, supplemental material uploaded as the separate pdf file

While it is commonly accepted that the disorder induced by magnetic ion doping in quantum magnets usually generates a rugged free-energy landscape resulting in slow or glassy spin dynamics, the disorder/distortion effects associated with non-magnetic ion sites doping are still illusive. Here, using AC susceptibility measurements, we show that the mixture of Sn/Ti on the non-magnetic ion sites of pyrochlore $Yb_2(Ti_{1-x}Sn_x)_2O_7$ induces an antiferromagnetic ground state despite both parent compounds, $Yb_2Ti_2O_7$, and $Yb_2Sn_2O_7$, order ferromagnetically. Local structure studies through neutron total scattering reveals the local distortion in the non-magnetic ion sites and its strong correlation with the magnetic phase switching. Our study, for the first time, demonstrates the local distortion as induced by the non-magnetic ion site mixture could be a new path to achieve magnetic phase switching, which has been traditionally obtained by external stimuli such as temperature, magnetic field, pressure, strain, light, etc.

The association of subordination and special functions is used to find sharp estimates on the parameter $\beta$ such that the analytic function $p(z)$ is subordinate to certain functions having positive real part whenever $p(z)+\beta z p'(z)$ is subordinate to the Janowski function. Further, when the traditional approach of solving higher order differential subordination implications failed, the concept of admissibility is employed to establish certain second and third order differential subordination relations between the analytic function $p$ and the functions associated with right half plane. As a sequel, we demonstrated the starlikeness of various well-known analytic functions as well.