A5019065183- Robotic Path Planning Algorithms
- Distributed Control Multi-Agent Systems
- Robotics and Sensor-Based Localization
- Autonomous Vehicle Technology and Safety
- Electron and X-Ray Spectroscopy Techniques
- Advanced Fluorescence Microscopy Techniques
- Adaptive Control of Nonlinear Systems
- Machine Learning and ELM
- Infrared Target Detection Methodologies
- Image Processing Techniques and Applications
- RNA Interference and Gene Delivery
- Domain Adaptation and Few-Shot Learning
- Target Tracking and Data Fusion in Sensor Networks
- Geophysical Methods and Applications
- Guidance and Control Systems
- SARS-CoV-2 and COVID-19 Research
- Graphene research and applications
- Advanced Memory and Neural Computing
- Software Reliability and Analysis Research
- Advanced Electron Microscopy Techniques and Applications
- UAV Applications and Optimization
- Advanced biosensing and bioanalysis techniques
- Lipid Membrane Structure and Behavior
- Topological Materials and Phenomena
- Immunotherapy and Immune Responses
Seoul National University
2020-2025
Institute for Basic Science
2024-2025
Advanced Institute of Convergence Technology
2024-2025
Seoul National University of Science and Technology
2024-2025
This letter presents a new online multi-agent trajectory planning algorithm that guarantees to generate safe, dynamically feasible trajectories in cluttered environment. The proposed utilizes linear safe corridor (LSC) formulate the distributed optimization problem with only constraints, so it does not resort slack variables or soft constraints avoid failure. We adopt priority-based goal method prevent deadlock without an additional procedure decide which robot yield. can compute for 60...
This letter presents a new distributed multi-agent trajectory planning algorithm that generates safe, dynamically feasible trajectories considering the uncertainty of obstacles in dynamic environments. We extend relative safe flight corridor (RSFC) presented previous work to replace time-variant, non-convex collision avoidance constraints convex ones, and we adopt relaxation method based on reciprocal (RCA) reduce total time distance without loss success rate. The proposed can compute for 50...
Abstract Owing to the exotic state of quantum matter, topological insulators have emerged as a significant platform for new‐generation functional devices. Among these insulators, tetradymites received attention because their van der Waals (vdW) structures and inversion symmetries. Although this symmetry completely blocks phenomena, it should be broken down facilitate versatile functionalities. Recently, Janus structure is suggested asymmetric out‐of‐plane lattice structures, terminating...
Decentralized multi-agent trajectory planning (MATP) can enhance the efficiency of multi-robot systems thanks to high scalability and short computation time. However, it may lead deadlock or livelock in obstacle-rich environments. To tackle this challenge, paper presents a decentralized MATP algorithm for quadrotor swarm that ensures convergence goal maze-like The proposed method guides agents their using waypoints generated by grid-based path (MAPF) algorithm. Additionally, we introduce...
Cross-Domain Few-Shot Learning (CDFSL) methods typically parameterize models with task-agnostic and task-specific parameters. To adapt parameters, recent approaches have utilized fixed optimization strategies, despite their potential sub-optimality across varying domains or target tasks. address this issue, we propose a novel adaptation mechanism called Task-Specific Preconditioned gradient descent (TSP). Our method first meta-learns Domain-Specific Preconditioners (DSPs) that capture the...
This article presents an online distributed trajectory planning algorithm for a quadrotor swarm in maze-like dynamic environment. We utilize linear safe corridor to construct the feasible collision constraints that can ensure interagent avoidance and consider uncertainty of moving obstacles. introduce mode-based subgoal resolve deadlock faster complex environment using only previously shared information. For obstacle avoidance, we adopt heuristic methods such as alert propagation escape...
Abstract Since the coronavirus pandemic, mRNA vaccines have revolutionized field of vaccinology. Lipid nanoparticles (LNPs) are proposed to enhance delivery efficiency; however, their design is suboptimal. Here, a rational method for designing LNPs explored, focusing on ionizable lipid composition and structural optimization using machine learning (ML) techniques. A total 213 analyzed random forest regression models trained with 314 features predict expression efficiency. The models, which...
<title>Abstract</title> Aberration correction is an important aspect of modern high-resolution scanning transmission electron microscopy. Most methods aligning aberration correctors require specialized sample regions and are unsuitable for fine-tuning aberrations without interrupting on-going experiments. Here, we present automated method correcting first- second-order called BEACON which uses Bayesian optimization the normalized image variance to efficiently determine optimal corrector...
Multi-robot systems (MRS) enable cooperation between multiple robots to achieve common goals or tasks. These can enhance efficiency and productivity in various applications, such as transportation, manufacturing, exploration. However, a critical issue MRS operation is the possibility of collisions with static/dynamic obstacles. This survey presents latest trends advancements collision avoidance approaches for multi-robot systems. We analyze centralized distributed methods, examining overall...
This letter presents a versatile trajectory planning pipeline for aerial tracking. The proposed tracker is capable of handling various chasing settings such as complex unstructured environments, crowded dynamic obstacles and multiple-target following. Among the entire pipeline, we focus on developing predictor future target motion planner. For rapid computation, employ sample-check-select strategy: modules sample set candidate movements, check multiple constraints, then select best...
Aberration correction is an important aspect of modern high-resolution scanning transmission electron microscopy. Most methods aligning aberration correctors require specialized sample regions and are unsuitable for fine-tuning aberrations without interrupting on-going experiments. Here, we present automated method correcting first- second-order called BEACON which uses Bayesian optimization the normalized image variance to efficiently determine optimal corrector settings. We demonstrate its...
Cross-Domain Few-Shot Learning~(CDFSL) methods typically parameterize models with task-agnostic and task-specific parameters. To adapt parameters, recent approaches have utilized fixed optimization strategies, despite their potential sub-optimality across varying domains or target tasks. address this issue, we propose a novel adaptation mechanism called Task-Specific Preconditioned gradient descent~(TSP). Our method first meta-learns Domain-Specific Preconditioners~(DSPs) that capture the...
The motion planning problem for multiple unstop-pable agents is of interest in many robotics applications, example, autonomous traffic management fixed-wing aircraft. Unfortunately, the existing algorithms cannot provide safety such agents, because they require to be able brake a complete stop and feasibility insurance. In this paper, we present distributed multi-agent planner that guarantees collision avoidance persistent feasibility, which can applied team homogeneous mobile vehicles stop....
In contrast to recent developments in online motion planning follow a single target with drone among obstacles, multi-target case chaser has been hardly discussed similar settings. Following more than one is challenged by multiple visibility issues due the inter-target occlusion and limited field-of-view addition possible collision obstacles. Also, reflecting targets into objectives or constraints increases computation load numerical optimization compared case. To resolve issues, we first...