IROS 2025 Papers — Page 5
IEEE/RSJ International Conference on Intelligent Robots and Systems · 1984 papers
Continuously Improved Reinforcement Learning for Automated Driving
Xuerun Yan, Haoran Wang
Autonomous DrivingReinforcement Learning
🎯 What it does: Proposed the High Confidence Policy Improvement (HCPI-RL) planner to achieve monotonic performance improvement in autonomous driving
Contrastive Autoencoder for Robust State Modelling of Soft Robots in Incomplete and Noisy Environments
Shageenderan Sapai, Chee Pin Tan
Robotic IntelligenceAuto EncoderContrastive LearningTime Series
🎯 What it does: Proposed the Contrastive Dual Latent Autoencoder (CDLAE), which uses an attention autoencoder and dual latent pathways to simultaneously handle missing data and noise in soft robot data, while employing contrastive loss to enforce separation between noise and clean signals; additionally, the autoencoder is jointly trained with a downstream prediction network to tightly couple signal interpolation with control tasks.
Control and Localization of Magnetic Nanorobot Swarms in Human-Sized Vascular Phantom
Shengyuan Wang, Lin Feng
Robotic IntelligenceBiomedical DataComputed Tomography
🎯 What it does: Study the control and spatial localization of magnetically controlled micro/nano-robot swarms in a highly realistic human-scale vascular model;
Control Marine Vehicles with Azimuth Thrusters using Convex Constrained Quadratic Programming
Mingxi Zhou, Chengzhi Yuan
OptimizationRobotic Intelligence
🎯 What it does: Propose a new method to reformulate the azimuth thruster control allocation problem as a convex quadratic programming problem, and verify its effectiveness on a simulated AUV.
Control the Soft Robot Arm with its "Physical Twin"
Qinghua Guan, Josie Hughes
Robotic Intelligence
🎯 What it does: A method for teleoperation using a physical twin identical to the soft robot is proposed, achieving full-body configuration perception by measuring tendon length and directly mapping it to actuators to control the movement of the soft robot arm;
Controllable Traffic Simulation through LLM-Guided Hierarchical Reasoning and Refinement
Zhiyuan Liu, Jianqiang Wang
Autonomous DrivingTransformerLarge Language ModelDiffusion model
🎯 What it does: Proposes a traffic simulation framework based on diffusion models and large language models (LLM), achieving hierarchical reasoning and self-reflection through a high-level understanding module and a low-level refinement module, while introducing a geometric quantization cost function based on the Frenet framework.
Convex Hull-based Algebraic Constraint for Visual Quadric SLAM
Xiaolong Yu, T. Feng
SegmentationPose EstimationOptimizationSimultaneous Localization and MappingImagePoint Cloud
🎯 What it does: Propose and apply convex hull-based algebraic constraints to object reconstruction, front-end pose estimation, and back-end bundle adjustment for quadratic curves.
ConViTac: Aligning Visual-Tactile Fusion with Contrastive Representations
Zhiyuan Wu, Shan Luo
Representation LearningTransformerContrastive LearningMultimodality
🎯 What it does: Proposed a visual-tactile fused representation learning network called ConViTac, which maps visual and tactile inputs to a unified latent embedding using a self-supervised contrastive learning pre-trained contrastive encoder, and achieves feature alignment through cross-modal attention;
ConvoyLLM: Dynamic Multi-Lane Convoy Control Using LLMs
Liping Lu, Pan Zhou
Autonomous DrivingOptimizationTransformerLarge Language Model
🎯 What it does: Proposed a dynamic multi-lane platoon control method based on large language models (LLMs), utilizing knowledge-driven real-time adaptive decision-making to achieve tasks such as obstacle avoidance, platoon joining/leaving, and formation switching, while ensuring the stability and flexibility of the overall platoon structure through an interleaved formation control strategy.
Cooperative Bearing-Only Target Pursuit via Multiagent Reinforcement Learning: Design and Experiment
Jianan Li, Shiyu Zhao
Reinforcement Learning
🎯 What it does: Studying the state estimation and tracking control problems for multi-robot systems tracking unknown targets
Cooperative Multi-Robot Path Finding with Removable Obstacles for Autonomous Environment Modification
Usha Kiruthika, Krupasagar Reddy
OptimizationRobotic Intelligence
🎯 What it does: A multi-robot path planning framework is proposed, enabling robots to remove movable obstacles (pits) after they are filled with sandbags, to achieve collaborative environment modification while minimizing total energy consumption.
CooperRisk: A Driving Risk Quantification Pipeline with Multi-Agent Cooperative Perception and Prediction
Mingyue Lei, Jiaqi Ma
Autonomous DrivingTransformerMultimodality
🎯 What it does: Proposes CooperRisk, an end-to-end pipeline that quantifies driving risk by leveraging multi-vehicle collaborative perception and prediction;
CooPre: Cooperative Pretraining for V2X Cooperative Perception
Seth Z. Zhao, Jiaqi Ma
Object DetectionAutonomous DrivingAuto EncoderPoint Cloud
🎯 What it does: Propose a self-supervised learning framework, CooPre, which leverages unlabeled 3D V2X data to aggregate multi-agent perception information and enhances collaborative perception performance through a point cloud reconstruction proxy task.
Coordinated Energy-Trajectory Economic Model Predictive Control for Autonomous Surface Vehicles under Disturbances*
Zhongqi Deng, Yaonan Wang
Autonomous DrivingOptimizationTime Series
🎯 What it does: Propose a coordinated energy-trajectory economic model predictive control (EMPC) scheme aimed at simultaneously optimizing ASV path tracking accuracy and energy consumption while achieving balance under environmental disturbances.
Coordination of Learned Decoupled Dual-Arm Tasks through Gaussian Belief Propagation
Adrián Prados, Ramón Barber
Robotic IntelligenceGraph Neural Network
🎯 What it does: Developed a dual-arm control algorithm based on learning from demonstrations, utilizing Gaussian Belief Propagation for task coordination
CoPAD : Multi-source Trajectory Fusion and Cooperative Trajectory Prediction with Anchor-oriented Decoder in V2X Scenarios
Kangyu Wu, Ya Zhang
Autonomous DrivingTransformerTime SeriesSequential
🎯 What it does: Propose the CoPAD framework to achieve the fusion of multi-source trajectory data and collaborative trajectory prediction.
CORENet: Cross-Modal 4D Radar Denoising Network with LiDAR Supervision for Autonomous Driving
Fuyang Liu, Yu Hu
RestorationAutonomous DrivingPoint Cloud
🎯 What it does: Propose CORENet, a 4D radar noise removal framework based on LiDAR supervision
Correspondence-Free Multiview Point Cloud Registration via Depth-Guided Joint Optimisation
Yiran Zhou, Liang Zhao
Pose EstimationOptimizationSimultaneous Localization and MappingPoint Cloud
🎯 What it does: Proposed a correspondence-free multi-view point cloud registration method by representing the global map as a depth map and constructing a nonlinear least squares optimization using original depth information to jointly estimate the point cloud pose and the global map.
Correspondence-Free Pose Estimation with Patterns: A Unified Approach for Multi-Dimensional Vision
Quan Quan, Dun Dai
Pose EstimationOptimization
🎯 What it does: Propose a new correspondence-free pose estimation method and its practical algorithm, with the core idea of eliminating unknowns through an additive process, separating pose estimation from correspondences, using point sets as patterns, and defining feature functions to generate sufficient equations for optimization.
Cost Function Estimation Using Inverse Reinforcement Learning with Minimal Observations
Sarmad Mehrdad, Ludovic Righetti
OptimizationReinforcement Learning
🎯 What it does: Propose an iterative inverse reinforcement learning algorithm to infer the optimal cost function in continuous space.
Crawler Robot with Movable Bending Point for Enhanced Traversability
Yuki Uda, Motoji Yamamoto
Robotic Intelligence
🎯 What it does: A crawling robot with a movable bending point was developed, capable of bending at any position along its body and dynamically adjusting the bending starting point via moving motor units to enhance its ability to navigate through narrow spaces, climb stairs, and cross gaps.
CRESSim–MPM: A Material Point Method Library for Surgical Soft Body Simulation with Cutting and Suturing
Yafei Ou, Mahdi Tavakoli
Physics Related
🎯 What it does: Designed and implemented CRESSim-MPM, a GPU-accelerated MPM library for surgical soft tissue physics simulation, supporting cutting and suturing.
Cross-Activity sEMG-Driven Joint Angle Estimation via Hybrid Attention Fusion: Bridging Traditional Features and Deep Spatial Representations
Zhimin Tang, Zhu Liang Yu
Pose EstimationConvolutional Neural NetworkRecurrent Neural NetworkTime SeriesBiomedical Data
🎯 What it does: Developed a HybridFusionAtt model based on surface electromyography (sEMG) for continuous joint angle estimation
Cross-Embodiment Robotic Manipulation Synthesis via Guided Demonstrations through CycleVAE and Human Behavior Transformer
Apan Dastider, Mingjie Lin
Robotic IntelligenceTransformerAuto Encoder
🎯 What it does: This paper proposes a method for cross-entity robot operation synthesis using CycleVAE and a human behavior Transformer. It aligns latent motion sequences across different entities through unsupervised CycleVAE and bidirectional subspace alignment, and employs a causal human behavior Transformer to learn the inherent motion dynamics from human expert demonstrations, ultimately generating smooth, complex task execution trajectories for robots.
CROSS-GAiT: Cross-Attention-Based Multimodal Representation Fusion for Parametric Gait Adaptation in Complex Terrains
Gershom Seneviratne, Dinesh Manocha
ClassificationRobotic IntelligenceConvolutional Neural NetworkTransformerMultimodality
🎯 What it does: Proposed an algorithm named CROSS-GAiT, which utilizes cross-attention to fuse visual and temporal inputs into terrain representations, enabling real-time adjustment of step height and hip expansion angle to adapt to complex terrains.
Cross-Level Fusion: Integrating Object Lists with Raw Sensor Data for 3D Object Tracking
Xiangzhong Liu, Hao Shen
Object TrackingAutonomous DrivingTransformerPoint Cloud
🎯 What it does: Propose a cross-layer fusion paradigm to achieve bidirectional information flow between object lists and raw visual features in 3D object detection and tracking, integrated into an end-to-end Transformer framework.
Cross-modal State Space Modeling for Real-time RGB-thermal Wild Scene Semantic Segmentation
Xiaodong Guo, Wujie Zhou
SegmentationConvolutional Neural NetworkMultimodality
🎯 What it does: Proposed the CM-SSM framework based on cross-modal state space modeling for real-time RGB-thermal pixel semantic segmentation in outdoor scenes.
CrossBEV-PR: Cross-modal Visual-LiDAR Place Recognition via BEV Feature Distillation
Jianbo Xu, Hesheng Wang
RecognitionAutonomous DrivingKnowledge DistillationImageMultimodalityPoint Cloud
🎯 What it does: Proposed a cross-modal visual-lidar localization method based on BEV feature distillation, achieving end-to-end cross-modal localization between surrounding images and point clouds for the first time.
Crouch Gait Recognition of Children with Cerebral Palsy Based on CNN-LSTM Hybrid Model
Junhang Liu, Xinyu Wu
RecognitionConvolutional Neural NetworkRecurrent Neural NetworkTime SeriesBiomedical Data
🎯 What it does: Data from four gait phases of children with cerebral palsy were collected using the Vicon 3D motion capture system, and a CNN-LSTM hybrid model was proposed for identifying crouch gait.
CrowdQuery: Density-Guided Query Module for Enhanced 2D and 3D Detection in Crowded Scenes
Marius Dähling, J. M. Zöllner
Object DetectionTransformerImagePoint Cloud
🎯 What it does: Proposes CrowdQuery, a density-guided query module designed to enhance 2D and 3D detectors, improving detection performance in crowded scenarios.
CRUISE: Cooperative Reconstruction and Editing in V2X Scenarios using Gaussian Splatting
Haoran Xu, Hao Zhao
Object DetectionObject TrackingData SynthesisAutonomous DrivingGaussian SplattingBenchmark
🎯 What it does: Built a reconstruction and synthesis framework for V2X driving environments, utilizing decomposed Gaussian splatting to achieve scene reconstruction and flexible editing, and rendering images from both vehicle and infrastructure perspectives for large-scale data augmentation.
CSC-MPPI: A Novel Constrained MPPI Framework with DBSCAN for Reliable Obstacle Avoidance*
Leesai Park, Sanghyun Kim
Autonomous DrivingOptimization
🎯 What it does: Propose CSC-MPPI, a variant of MPPI with strong constraints on state and control inputs, to enhance trajectory optimization;
CSVO: Complementary-Pathway Spatial-Enhanced Visual Odometry for Extreme Environments with Brain-Inspired Vision Sensors
Yihan Lin, Rong Zhao
Pose EstimationAutonomous DrivingConvolutional Neural NetworkSimultaneous Localization and MappingImageVideoMultimodality
🎯 What it does: Proposed and implemented a CSVO visual odometry method based on the brain-inspired visual sensor Tianmouc, enhancing pose estimation in extreme environments by fusing information from the cognitive channel (COP) and action channel (AOP).
CTSG: Integrating Context and Way Topology Into Scene Graph for Zero-shot Navigation
Ruifei Ma, Chao Zhang
Autonomous DrivingGraph Neural NetworkVision-Language-Action ModelMultimodality
🎯 What it does: Proposes CTSG—a hierarchical 3D scene graph mapping framework that integrates environmental context and path topology for zero-shot target navigation, supporting both visual and text queries.
CueLearner: Bootstrapping and local policy adaptation from relative feedback
Giulio Schiavi, Roland Siegwart
Reinforcement Learning from Human FeedbackReinforcement Learning
🎯 What it does: Proposed the CueLearner method, integrating relative feedback with offline reinforcement learning to guide exploration and achieve policy adaptation.
Cumulative Informative Path Planning for Efficient Gas Source Localization with Mobile Robots
Wanting Jin, A. Martinoli
OptimizationComputational EfficiencyRobotic IntelligenceSimultaneous Localization and Mapping
🎯 What it does: Proposed an accumulative information path planning algorithm, achieving gas source localization without stopping using a perception-driven following strategy and lightweight information extraction metrics.
CushionCatch: A Compliant Catching Mechanism for Mobile Manipulators via Combined Optimization and Learning
Bingjie Chen, Bin Liang
OptimizationRobotic IntelligenceRecurrent Neural Network
🎯 What it does: Proposed a framework combining optimization and learning to achieve compliant grasping on a mobile manipulator
CVD-SfM: A Cross-View Deep Front-end Structure-from-Motion System for Sparse Localization in Multi-Altitude Scenes
Yaxuan Li, Brendan Englot
Pose EstimationTransformerImage
🎯 What it does: Propose a multi-height camera pose estimation system that achieves accurate localization using sparse images
CVIRO: A Consistent and Tightly-Coupled Visual-Inertial-Ranging Odometry on Lie Groups
Yizhi Zhou, Xuan Wang
Pose EstimationRobotic IntelligenceSimultaneous Localization and MappingMultimodality
🎯 What it does: Proposed the CVIRO system based on Lie group theory, which integrates UWB, visual, and inertial information to achieve joint consistency estimation of robot and UWB anchor states.
CVLN-Think: Causal Inference with Counterfactual Style Adaptation for Continuous Vision-and-Language Navigation
Ruonan Liu, Weidong Zhang
Autonomous DrivingVision-Language-Action Model
🎯 What it does: Proposed the CVLN-Think (CVT) model, enhancing the robustness and adaptability of continuous vision-language navigation (VLN-CE) through causal inference. The model includes a Style Causal Adapter (SCA) and a Thinking Causal Navigation Engine (TCNE), enabling agents to learn invariant spatial structures and actively adjust navigation decisions.
D4orm: Multi-Robot Trajectories with Dynamics-aware Diffusion Denoised Deformations
Yuhao Zhang, Amanda Prorok
OptimizationRobotic IntelligenceDiffusion model
🎯 What it does: Proposed an incremental denoising approach using diffusion models to generate feasible and collision-free multi-robot trajectories.
DA-MPPI: Disturbance-Aware Model Predictive Path Integral via active disturbance estimation and compensation
Haodi Zhang, Shihua Li
Autonomous DrivingOptimization
🎯 What it does: By integrating the Extended High-Order Sliding Mode Observer (ESMO) into MPPI, active estimation and compensation of external disturbances were achieved, thereby improving trajectory tracking performance.
DarkSeg: Infrared-Driven Semantic Segmentation for Garment Grasping Detection in Low-Light Conditions
Haifeng Zhong, Yixing Gao
SegmentationDepth EstimationKnowledge DistillationRobotic IntelligenceImageMultimodality
🎯 What it does: Designed and implemented a low-light clothing grasping detection model called DarkSeg, utilizing a student-teacher model for feature alignment, learning illumination-invariant structural representations from an infrared teacher model, and proposing a depth-aware grasping strategy while constructing the DarkClothes dataset;
dARt Vinci: Egocentric Data Collection for Surgical Robot Learning at Scale
Yihao Liu, Mehran Armand
Data-Centric LearningRobotic IntelligenceWorld ModelVideoBiomedical Data
🎯 What it does: Proposed dARt Vinci, a scalable, view-centered surgical robot data collection platform constructed using augmented reality gesture tracking and high-fidelity physics engine;
DashGaze: Driver Gaze Through Dashcam
T. John, C. V. Jawahar
Pose EstimationAutonomous DrivingConvolutional Neural NetworkImageVideoBenchmark
🎯 What it does: Collected driver gaze data using in-vehicle cameras and constructed the DashGaze large-scale synchronized perspective dataset along with the baseline model DashGazeNet.
Data-Bootstrapped, Physics-Informed Framework for Object Rearrangement
Alex Wong, Zhiwei Dong
Robotic IntelligenceTransformerReinforcement LearningPhysics Related
🎯 What it does: Proposed a data bootstrapping and physics-informed object reordering framework called DPR.
Data-Driven Fault Detection for Wafer Scanner Cable Slabs using Koopman Operators
Michael Pumphrey, M. Al Janaideh
Anomaly DetectionTime Series
🎯 What it does: Propose a data-driven early fault diagnosis framework
Data-Driven MPC for Attitude Control of Autonomous Underwater Robot
Tianzhu Gao, Yantao Shen
OptimizationRobotic IntelligenceTime Series
🎯 What it does: Model the dynamics of underwater robots using a data-driven SINDy method, and embed the learned model into a model predictive controller (MPC) to achieve attitude control of omni-directional propulsion underwater robots.
Data-driven Visual Servoing of Flexible Continuum Robots in Constrained Environments
Wei Chen, Yun-Hui Liu
Robotic Intelligence
🎯 What it does: Propose a data-driven, model-free control strategy for flexible continuous robots, achieving visual servoing through Lie bracket approximation.
DB-MPO: Demonstration Boosted Reactive Grasping For Two-Finger Gripper
Boya Zhang, G. Martius
Robotic IntelligenceReinforcement Learning from Human FeedbackReinforcement Learning
🎯 What it does: Proposes an offline actor-critic reinforcement learning method based on demonstration injection for interaction and policy optimization in two-finger grasping tasks
DBaS-Log-MPPI: Efficient and Safe Trajectory Optimization via Barrier States
Fanxin Wang, Yikun Cheng
Autonomous DrivingOptimizationSafty and Privacy
🎯 What it does: Proposed and verified a novel trajectory optimization algorithm called DBaS-Log-MPPI, which can efficiently solve paths and ensure safe navigation in nonlinear control systems;
DCM Modulation: A Three-Axis Rotation Stabilization Technique for Bipedal Locomotion Control
Yuichi Tazaki
Robotic Intelligence
🎯 What it does: Propose a simple bipedal gait controller that stabilizes roll, pitch, and yaw rotations without relying on ground reaction torque
DCT-Diffusion: Depth Completion for Transparent Objects with Diffusion Denoising Approach
Zhenning Zhou, Qixin Cao
Depth EstimationTransformerDiffusion model
🎯 What it does: Proposes the DCT-Diffusion framework for depth completion of transparent objects;
DecARt Leg: Design and Evaluation of a Novel Humanoid Robot Leg with Decoupled Actuation for Agile Locomotion
Egor Davydenko, Roman Gorbachev
Robotic Intelligence
🎯 What it does: Proposed and evaluated a new electric robot leg called DecARt Leg to achieve agile gaits; designed a new descriptive metric called FAST and conducted quantitative comparisons with other leg designs; validated its performance through extensive simulations and preliminary hardware experiments.
Decentralized admittance control for a multi-manipulator system: implementation and analysis
Graziano Carriero, Y. Karayiannidis
OptimizationRobotic Intelligence
🎯 What it does: Propose a decentralized strategy for multi-robot arm cooperative transportation of objects.
Decentralized but Not Compromised: Modular Architecture with Refined Observation for Multi-Agent Model-Based Reinforcement Learning
Shuqi Wang, Ping Wei
Reinforcement LearningWorld ModelSequential
🎯 What it does: Proposes a modular architecture named MARO, achieving the CTDE principle in multi-agent model-driven reinforcement learning through refined observations.
Decentralized Declustering of Multiple Underactuated Autonomous Surface Vehicles: Managing Robot Swarms in the Field
Filip Traasdahl Strømstad, Michael Benjamin
OptimizationRobotic Intelligence
🎯 What it does: Propose a decentralized and scalable method to transition multiple underactuated autonomous surface vehicles from arbitrary initial states to conflict-free deployment states.
Decentralized Gaussian Process Classification and an Application in Subsea Robotics
Yifei Gao, James McMahon
ClassificationRobotic IntelligenceAudio
🎯 What it does: Developed a decentralized classification framework for real-time communication success probability mapping and proposed a data sharing strategy.
Decentralized Model-Free Monitoring of Multi-UAV-Multi-USV Systems Using Sparse Data and Bayesian Learning
Jiajie Huang, Hai-Tao Zhang
Optimization
🎯 What it does: Propose a method utilizing multiple UAVs through sparse Bayesian learning and Kalman filtering for model-free, decentralized monitoring and trajectory optimization of multi-USV systems
Decentralized Multi-robot Navigation Policy with Enhanced Security Using Graph GRU Policy Network
Lin Chen, Danwei Wang
Robotic IntelligenceRecurrent Neural NetworkGraph Neural NetworkReinforcement Learning
🎯 What it does: Proposed a decentralized multi-robot navigation strategy based on graph attention + GRU, and applied it in a simulation environment
Decentralized Uncertainty-Aware Multi-Agent Collision Avoidance With Model Predictive Path Integral*
S. Dergachev, Konstantin S. Yakovlev
Autonomous DrivingOptimizationSafty and Privacy
🎯 What it does: Proposed a decentralized multi-agent collision avoidance method that integrates Model Predictive Path Integral (MPPI) with probability-adapted Optimal Reciprocal Collision Avoidance, and directly embeds probabilistic safety constraints into the MPPI sampling process through second-order cone programming (SOCP);
Decision Transformer-Based Drone Trajectory Planning with Dynamic Safety–Efficiency Trade-Offs
Chang-Hun Ji, Sungtae Moon
Autonomous DrivingOptimizationTransformerReinforcement Learning
🎯 What it does: Propose a UAV trajectory planner based on the Decision Transformer, utilizing the Return-to-Go (RTG) parameter to dynamically adjust the trade-off between safety and efficiency, and validate it in Gazebo simulation and real-world environments.
Decremental Dynamics Planning for Robot Navigation
Yuanjie Lu, Xuesu Xiao
Robotic Intelligence
🎯 What it does: Proposed and implemented a decremental dynamic programming (DDP) paradigm that gradually reduces robot dynamics constraints from high-fidelity models to low-fidelity models to bridge the gap between global and local planning, applying it to three different planners and a newly developed navigation system; the system achieved second place in both the simulation and physical testing phases of the 2025 BARN challenge.
Deep Coarse-to-Fine Networks for Robust Segmentation and Pose Estimation of Surgical Suturing Threads
Xinyao Zhou, Guang-Zhong Yang
SegmentationPose EstimationBiomedical Data
🎯 What it does: Proposed a coarse-to-fine network for fine-grained segmentation and suture pose estimation.
Deep Equivariant Multi-Agent Control Barrier Functions
N. Bousias, G. Pappas
Graph Neural Network
🎯 What it does: Proposed a symmetry-integrated distributed control barrier function for safe control in multi-agent systems.
Deep Learning-Based Pig Behavior Captioning for Smart Livestock Farming
Honghua Jiang, Yongliang Qiao
GenerationConvolutional Neural NetworkRecurrent Neural NetworkGraph Neural NetworkVision Language ModelImageMultimodalityAgriculture Related
🎯 What it does: Propose a multimodal image description model to generate semantic text descriptions of pig behaviors, supporting smart farm decision-making.
Deep Learning-based Proactive Hazard Prediction for Human-Robot Collaboration with Sensor Malfunctions
Yuliang Ma, Andrey Morozov
Anomaly DetectionSafty and PrivacyRobotic IntelligenceMultimodality
🎯 What it does: Propose an active danger prediction method based on deep learning, which can predict potential dangers in human-robot collaboration after detecting sensor anomalies.
Deep Predictive Learning with Proprioceptive and Visual Attention for Humanoid Robot Repositioning Assistance
Tamon Miyake, Shigeki Sugano
Robotic IntelligenceTransformerImage
🎯 What it does: Utilizing a deep neural network combined with proprioception and visual attention mechanisms to implement posture relocalization-assisted actions on the dual-arm humanoid robot Dry-AIREC
Deep Reinforcement Learning-Based Levitation Control of Wireless Capsule Endoscope by Robotically Driven Permanent Magnet
Ding Huang, Chengzhi Hu
Robotic IntelligenceRecurrent Neural NetworkReinforcement Learning
🎯 What it does: Developed a magnetic suspension control method based on permanent magnets, utilizing deep reinforcement learning to achieve five-degree-of-freedom navigation of wireless capsule endoscopes in a simulated environment.
Deep Reinforcement Learning-Based Trajectory Tracking Framework for 4WS Robots Considering Switch of Steering Modes
Runjiao Bao, Tianwei Niu
Robotic IntelligenceReinforcement Learning
🎯 What it does: Proposed a multi-modal trajectory tracking method considering steering mode switching, dividing the trajectory tracking task into mode decision and tracking control, and implementing it based on deep reinforcement learning;
DEEP-SEA: Deep-Learning Enhancement for Environmental Perception in Submerged Aquatics
Shuang Chen, Amir Atapour-Abarghouei
RestorationTransformerImage
🎯 What it does: Proposed the DEEP-SEA deep learning model to enhance both low-frequency and high-frequency information in underwater images while preserving spatial structure.
Delving into Mapping Uncertainty for Mapless Trajectory Prediction
Zongzheng Zhang, Hao Zhao
Autonomous DrivingSimultaneous Localization and MappingPoint Cloud
🎯 What it does: Analyze the impact of map uncertainty in mapless trajectory prediction, and propose Proprioceptive Scenario Gating and Covariance Map Uncertainty methods to adaptively integrate map uncertainty into trajectory prediction models.
Demonstration-Enhanced Adaptable Multi-Objective Robot Navigation
Jorge de Heuvel, Maren Bennewitz
Robotic IntelligenceReinforcement Learning
🎯 What it does: Proposes a framework combining demonstration learning with multi-objective reinforcement learning (MORL), enabling robots to dynamically adapt without retraining when user preferences change.
Dense Semantic Bird-Eye-View Map Generation from Sparse LiDAR Point Clouds via Distribution-aware Feature Fusion
Jinsong Li, Yuxiang Sun
GenerationAutonomous DrivingPoint Cloud
🎯 What it does: Proposes the PointDenseBEV framework, which can directly generate dense semantic bird's-eye-view (BEV) maps from sparse LiDAR point clouds.
Depth Estimation Based on Fisheye Cameras
Yuwei Zhou, Guoyu Lu
Depth EstimationConvolutional Neural NetworkImage
🎯 What it does: This paper proposes a deep learning-based depth estimation framework for fisheye cameras, which is trained using calibrated synchronized stereo image pairs. It combines SSIM and L1 spatial consistency loss, and improves reconstruction accuracy by introducing uncertainty maps and distortion distribution maps through a fisheye-specific depth refinement module. During inference, accurate depth maps can be generated from a single fisheye image.
Depth Matters: Exploring Deep Interactions of RGB-D for Semantic Segmentation in Traffic Scenes
Siyu Chen, Guorong Cai
SegmentationAutonomous DrivingTransformerImageMultimodality
🎯 What it does: Propose a learnable Depth Interaction Pyramid Transformer (DiPFormer) that leverages depth information to enhance semantic segmentation and road detection in traffic scenarios using RGB-D data.
Design and Active Stability Control of a Wheel-foot Mobile Platform with High Trafficability
Xiran Li, Han Yuan
Robotic Intelligence
🎯 What it does: Designed and built a mobile platform with a wheel-legged hybrid mechanism, implemented an adaptive planning and active stabilization control system, and completed dynamic modeling and experimental verification.
Design and Characterization of a Thermal-electrostatic Dual-modal Soft Pouch Motor
Chuang Wu, Chongjing Cao
Robotic Intelligence
🎯 What it does: Designed and characterized a thermoelectric-static dual-mode soft membrane motor, and used it in series to construct a folding fan-style actuator and an accordion-style soft gripper, demonstrating potential applications in soft robotics.
Design and Control of a 6-DOF Fully Actuated Aerial-Aquatic Robot with Thrust Vectoring
Bocheng Tian, Li Wen
OptimizationRobotic Intelligence
🎯 What it does: Designed and controlled a six-degree-of-freedom, fully actuated air-water amphibious robot using thrust vectoring technology.
Design and Control of an Actively Morphing Quadrotor with Vertically Foldable Arms
Tingyu Yeh, Lijun Han
Robotic Intelligence
🎯 What it does: Proposes a quadrotor drone design with vertically foldable arms that can fly like traditional quadrotors and contract into a gripper to grasp objects and pass through narrow spaces.
Design and Control of SeparaTrek: A Hybrid Aerial-Ground Robot with Separable and Combinative Locomotion Parts
Yu Zhang, Yanbo Sun
Robotic Intelligence
🎯 What it does: Designed a hybrid aerial-ground robot named SeparaTrek with separable and combinable structures to reduce aerial-ground functional coupling, enabling multi-mode ground and aerial locomotion.
Design and Control of Soft Robotic Wearable with SMA-based Artificial Muscle Fibers for Ankle Assistance
Eunsung Joo, Je-sung Koh
Robotic Intelligence
🎯 What it does: Developed a garment-based soft wearable device with embedded SMA artificial muscle fibers for ankle plantar flexion assistance, achieving precise force and displacement control through closed-loop PI control.
Design and Development of a Deformable Spherical Robot for Amphibious Applications*
Le Xu, Li-sha Lu
Robotic Intelligence
🎯 What it does: Designed and developed a deformable spherical robot with a six-pillar topology structure, capable of achieving multi-mode motion in complex aquatic and terrestrial environments.
Design and Development of a GPR-Equipped Robot for Full-space External Diseases Detection in Drainage Pipelines*
Yuanjin Fang, Maoxuan Xu
Anomaly DetectionRobotic Intelligence
🎯 What it does: Designed and developed a pipeline robot equipped with ground-penetrating radar (GPR) capable of comprehensively detecting external space diseases in pipes with diameters ranging from 500 to 1000 millimeters.
Design and Development of a Propulsion Induced Rolling Spherical Tensegrity Robot*
Niansong Zhang, Jie Zhao
Robotic IntelligencePhysics Related
🎯 What it does: Designed and built a spherical tension structure robot driven by six fixed pneumatic thrusters, and conducted motion experiments in simulation and real environments.
Design and Development of the MeCO Open-Source Autonomous Underwater Vehicle
David Widhalm, Junaed Sattar
Computational EfficiencyRobotic IntelligenceImageAudio
🎯 What it does: This paper introduces MeCO, a medium-cost, open-source underwater unmanned vehicle designed to support underwater human-robot interaction and marine robotics research; the platform features economical, scalable, modular hardware and software designs, along with specialized systems for human-robot interaction, including front and rear displays, optical communication devices, acoustic interaction transducers, and stereo vision.
Design and Dynamic Modeling Analysis of Undulatory Propulsion Underwater Robot with Rotational Passive Degrees of Freedom in Fin Rays *
Tangjia Zhang, Liangjie Sun
Physics Related
🎯 What it does: Designed and studied a whip-like fin underwater robot with a rotating passive joint, established a model of fin oscillation and passive rotational degrees of freedom, a hydrodynamic model based on fluid drag theory, and a full dynamic model, and verified its performance through numerical simulation and experimental testing.
Design and Flight Control of a Novel Thrust-Vectored Tricopter Using Twisting and Tilting Rotors
Xinliang Li, Kun Liu
Robotic Intelligence
🎯 What it does: A novel, compact over-driven trirotor was designed and implemented, employing servo-driven torsion and tilt mechanisms to avoid internal force conflicts during flight, and an improved force decomposition (FD)-based control allocation method was proposed, considering gyroscopic torques generated by arm rotation. The dynamic model of the rotating joints was utilized as a virtual sensor to estimate the improved control effectiveness matrix, and its performance was validated through simulation and flight experiments.
Design and Geometry-Aware Planning of a Novel Probe-Scanning Manipulator with RCM Constraint
Xiao Luo, Zheng Li
OptimizationRobotic IntelligenceUltrasound
🎯 What it does: Designed and developed a six-degree-of-freedom parallel-serial hybrid structure TRUS probe scanning manipulator that satisfies the remote center of motion (RCM) constraint, and provided its kinematic model; proposed a geometric-aware path planning method based on SO(3) Riemannian geometry and a smooth rotational trajectory generation method that minimizes angular acceleration; validated the system's effectiveness and practicality through simulation and experiments.
Design and Implement of Large-scale Tail-sitter VTOL UAV
Zhixiong Xu, Li Fan
OptimizationRobotic IntelligencePhysics Related
🎯 What it does: Designed and implemented a large-scale tail-sitting VTOL drone, adopting an H-shaped structure with forward-swept wings to enhance longitudinal stability. Structural optimization balanced strength and lightweight design, while the drive system utilized four motors and eight servos. Computational fluid dynamics (CFD) simulations and wind tunnel experiments were conducted to characterize aerodynamics across all conditions, and a control framework covering the entire flight envelope was developed. Finally, performance and reliability were validated through simulation and outdoor flight tests.
Design and Integration of an Optical Frequency Domain Reflectometry (OFDR) Sensor with a Flexible Pedicle Screw for Biomechanical Evaluation
Y. Kulkarni, F. Alambeigi
Biomedical Data
🎯 What it does: Designed, integrated, calibrated, and evaluated a system for integrating optical frequency domain reflectometry (OFDR) strain sensors into flexible pedicle screws (FPS), and assessed its shape sensing performance through static and dynamic insertion experiments.
Design and Kinematics for the Cystoscope of a Transurethral Continuum Surgical Robotic System
Haomin Kuang, Y. H. Liu
Robotic IntelligenceBiomedical Data
🎯 What it does: Proposed an inverse kinematics solution algorithm for a continuous cystoscope used in TURBT
Design and Modeling of a Micro-coil Array Platform for the Smooth Movement of Multiple Micro-robots
Xinzhe Tang, Qigao Fan
Robotic IntelligencePhysics Related
🎯 What it does: A micro-coil array driving platform was proposed, and a differential current driving strategy was designed to achieve independent control of smooth motion for multiple micro-robots; through modeling of spatial magnetic field distribution, formulation of driving strategies based on magnetic force models, and construction of an experimental platform, a series of experiments were conducted for verification.
Design and Performance Analysis of a Pipeline Crawling Robot Based on Spring-Roll Dielectric Elastomer Actuators
Qinghai Zhang, Shijie Guo
OptimizationRobotic IntelligencePhysics Related
🎯 What it does: Designed a lightweight pipeline-crawling robot based on spring-wound dielectric elastomer actuators.
Design and Performance Study of an Underwater Soft Snake-like Robot
Huichen Ma, R. C. Yeow
Robotic Intelligence
🎯 What it does: Proposed a two-actuator underwater soft snake robot prototype made of 3D-printed soft materials, which generates sinusoidal bending motion by regulating actuators through control signals; experimental studies on tail material, phase shift, and voltage growth rate were conducted to observe their effects on motion performance.
Design and Wrench-Feasible-Workspace Analysis for a Novel Cable-driven Parallel Robot with Movable Anchor Winches
Hao An, Han Yuan
Robotic Intelligence
🎯 What it does: Designed and analyzed a spatially mobile cable-driven parallel robot (M-CDPR) with a cable winch capable of freely moving along a closed circular track, proposed a reconfiguration method for the cable winch, and validated its feasibility through simulation and prototype experiments.
Design of a Module for a Modular Snake Robot with 3D Locomotion
Yuya Shimizu, T. Kamegawa
Robotic Intelligence
🎯 What it does: A modular snake robot composed of a minimal module (5 links and 4 joints) was proposed, which can rotate and move in a 2D plane. Each module is equipped with magnetic and mechanical hooks to achieve autonomous separation and docking without human intervention; field experiments verified the repeatability of the separation/docking actions and the mobility of a single module in the separated state.
Design of a Six-Bar Linkage-Inspired Reversible Wing for Stopped-Rotor Vehicles
Kristan Hilby, Ian W. Hunter
Optimization
🎯 What it does: A six-bar linkage-driven reversible deformable wing was designed and studied. This wing combines flexibility with rigidity against aerodynamic loads. The structural performance was verified using a one-way fluid-structure coupling simulation, and aerodynamic efficiency was improved through constrained optimization.
Design of a Soft Automatic Anchoring System for Enhanced Mobility and Stability in Colonoscopy Robots
Yiying Liang, Xin Ma
Robotic Intelligence
🎯 What it does: Propose a soft automatic anchoring system (SAAS) to enhance the mobility and stability of a colonoscopy robot within the colon.
Design of a swimming microrobot powered by a single piezoelectric bender
Cameron Urban, E. F. Helbling
Robotic IntelligencePhysics Related
🎯 What it does: Designed, manufactured, and assembled a millimeter-scale fish-like microrobot (Daniobot), and measured its displacement and velocity under different tail amplitudes and frequencies.