Deep neural networks (DNNs) can deteriorate in accuracy when deployment data differs from training data. While performing online training at all timesteps can improve accuracy, it is computationally expensive. We propose DecTrain, a new algorithm that decides when to train a monocular depth DNN online using self-supervision with low overhead. To make the decision at each timestep, DecTrain compares the cost of training with the predicted accuracy gain. We evaluate DecTrain on out-of-distribution data, and find DecTrain maintains accuracy compared to online training at all timesteps, while training only 44% of the time on average. We also compare the recovery of a low inference cost DNN using DecTrain and a more generalizable high inference cost DNN on various sequences. DecTrain recovers the majority (97%) of the accuracy gain of online training at all timesteps while reducing computation compared to the high inference cost DNN which recovers only 66%. With an even smaller DNN, we achieve 89% recovery while reducing computation by 56%. DecTrain enables low-cost online training for a smaller DNN to have competitive accuracy with a larger, more generalizable DNN at a lower overall computational cost.
{"title":"DecTrain: Deciding When to Train a Monocular Depth DNN Online","authors":"Zih-Sing Fu;Soumya Sudhakar;Sertac Karaman;Vivienne Sze","doi":"10.1109/LRA.2025.3536206","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536206","url":null,"abstract":"Deep neural networks (DNNs) can deteriorate in accuracy when deployment data differs from training data. While performing online training at all timesteps can improve accuracy, it is computationally expensive. We propose DecTrain, a new algorithm that decides when to train a monocular depth DNN online using self-supervision with low overhead. To make the decision at each timestep, DecTrain compares the cost of training with the predicted accuracy gain. We evaluate DecTrain on out-of-distribution data, and find DecTrain maintains accuracy compared to online training at all timesteps, while training only 44% of the time on average. We also compare the recovery of a low inference cost DNN using DecTrain and a more generalizable high inference cost DNN on various sequences. DecTrain recovers the majority (97%) of the accuracy gain of online training at all timesteps while reducing computation compared to the high inference cost DNN which recovers only 66%. With an even smaller DNN, we achieve 89% recovery while reducing computation by 56%. DecTrain enables low-cost online training for a smaller DNN to have competitive accuracy with a larger, more generalizable DNN at a lower overall computational cost.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2822-2829"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143396277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3536297
Mitchell Torok;Mohammad Deghat;Yang Song
Sampling-based model predictive control algorithms can be computationally expensive and may not be feasible for restricted platforms such as quadcopters. Comparatively speaking, lightweight learned controllers are computationally cheaper and may be more suited for these platforms. Expert control samples provided by a remote model predictive control algorithm could be used to rapidly train a student policy. We present Greedy-DAgger, a hybrid-policy imitation learning approach that leverages expert simulations to improve the student rollout efficiency during the training of a student policy. Our approach builds on the DAgger algorithm by employing a greedy strategy, that selects isolated states from a student trajectory. These states are used to generate expert trajectory samples before supervised learning is performed and the process is repeated. The effectiveness of the Greedy-DAgger algorithm is evaluated on two simulated robotic systems: a cart pole and a quadcopter. For these environments, Greedy-DAgger was shown to be up to ten times more rollout efficient than conventional DAgger. The introduced improvements enable expert-level quadcopter control to be achieved within 8 seconds of wall time. The Crazyflie quadcopter platform was then utilised to validate the simulation results and demonstrate the potential for real-world training with Greedy-DAgger on a constrained platform, leveraging access to a remote GPU-accelerated server.
{"title":"Greedy-DAgger - A Student Rollout Efficient Imitation Learning Algorithm","authors":"Mitchell Torok;Mohammad Deghat;Yang Song","doi":"10.1109/LRA.2025.3536297","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536297","url":null,"abstract":"Sampling-based model predictive control algorithms can be computationally expensive and may not be feasible for restricted platforms such as quadcopters. Comparatively speaking, lightweight learned controllers are computationally cheaper and may be more suited for these platforms. Expert control samples provided by a remote model predictive control algorithm could be used to rapidly train a student policy. We present Greedy-DAgger, a hybrid-policy imitation learning approach that leverages expert simulations to improve the student rollout efficiency during the training of a student policy. Our approach builds on the DAgger algorithm by employing a greedy strategy, that selects isolated states from a student trajectory. These states are used to generate expert trajectory samples before supervised learning is performed and the process is repeated. The effectiveness of the Greedy-DAgger algorithm is evaluated on two simulated robotic systems: a cart pole and a quadcopter. For these environments, Greedy-DAgger was shown to be up to ten times more rollout efficient than conventional DAgger. The introduced improvements enable expert-level quadcopter control to be achieved within 8 seconds of wall time. The Crazyflie quadcopter platform was then utilised to validate the simulation results and demonstrate the potential for real-world training with Greedy-DAgger on a constrained platform, leveraging access to a remote GPU-accelerated server.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2878-2885"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143403970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3536291
Zechang Wang;Dengpeng Xing;Yiming Yang;Peng Wang
Reinforcement learning (RL) typically presupposes instantaneous agent-environment interactions, but in real-world scenarios such as robotic control, overlooking observation delays can significantly impair performance. While existing studies consider stationary, known delays, real-world applications frequently encounter unpredictable delay variations. To address this problem, this letter presents a novel algorithm for scheduling delayed dynamic models. Specifically, We propose using multiple truncated delay distributions to effectively model time-varying delays, with each distribution tailored to learn a specific delayed dynamic model. These models map delayed observations and historical actions to the current state, integrating seamlessly with existing RL algorithms to facilitate optimal decision-making. Since the delay is unknown to the agent, we propose an effective delay estimation method to detect delay and their corresponding distributions in real-time, thereby adaptively selecting the most appropriate delayed dynamic model to manage delays. To reduce instability caused by abrupt changes in delay distribution and enhance responsiveness to such variations, we apply Bayesian online changepoint detection to enable rapid sensing of alterations in the delay distribution within a finite number of time-steps. To the best of our knowledge, our approach is the first effective solution to the non-stationary time-varying delay problem in RL. Empirical results demonstrate the robust performance of our method in scenarios characterized by non-stationary observation delays, highlighting its strong potential for robotic control applications.
{"title":"Delayed Dynamic Model Scheduled Reinforcement Learning With Time-Varying Delays for Robotic Control","authors":"Zechang Wang;Dengpeng Xing;Yiming Yang;Peng Wang","doi":"10.1109/LRA.2025.3536291","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536291","url":null,"abstract":"Reinforcement learning (RL) typically presupposes instantaneous agent-environment interactions, but in real-world scenarios such as robotic control, overlooking observation delays can significantly impair performance. While existing studies consider stationary, known delays, real-world applications frequently encounter unpredictable delay variations. To address this problem, this letter presents a novel algorithm for scheduling delayed dynamic models. Specifically, We propose using multiple truncated delay distributions to effectively model time-varying delays, with each distribution tailored to learn a specific delayed dynamic model. These models map delayed observations and historical actions to the current state, integrating seamlessly with existing RL algorithms to facilitate optimal decision-making. Since the delay is unknown to the agent, we propose an effective delay estimation method to detect delay and their corresponding distributions in real-time, thereby adaptively selecting the most appropriate delayed dynamic model to manage delays. To reduce instability caused by abrupt changes in delay distribution and enhance responsiveness to such variations, we apply Bayesian online changepoint detection to enable rapid sensing of alterations in the delay distribution within a finite number of time-steps. To the best of our knowledge, our approach is the first effective solution to the non-stationary time-varying delay problem in RL. Empirical results demonstrate the robust performance of our method in scenarios characterized by non-stationary observation delays, highlighting its strong potential for robotic control applications.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2646-2653"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143361295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Visual inspection of confined spaces such as aircraft wings is ergonomically challenging for human mechanics. This work presents a novel crane robot that can travel the entire span of the aircraft wing, enabling mechanics to perform inspection from outside of the confined space. However, teleoperation of the crane robot can still be a challenge due to the need to avoid obstacles in the workspace and potential oscillations of the camera payload. The main contribution of this work is to exploit the differential flatness of the crane-robot dynamics for designing reduced-oscillation, collision-free time trajectories of the camera payload for use in teleoperation. Autonomous experiments verify the efficacy of removing undesired oscillations by 89%. Furthermore, teleoperation experiments demonstrate that the controller eliminated collisions (from 33% to 0%) when 12 participants performed an inspection task with the use of proposed trajectory selection when compared to the case without it. Moreover, even discounting the failures due to collisions, the proposed approach improved task efficiency by 18.7% when compared to the case without it.
{"title":"Semi-Autonomous Teleoperation Using Differential Flatness of a Crane Robot for Aircraft In-Wing Inspection","authors":"Wade Marquette;Kyle Schultz;Vamsi Jonnalagadda;Benjamin Wong;Joseph Garbini;Santosh Devasia","doi":"10.1109/LRA.2025.3536207","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536207","url":null,"abstract":"Visual inspection of confined spaces such as aircraft wings is ergonomically challenging for human mechanics. This work presents a novel crane robot that can travel the entire span of the aircraft wing, enabling mechanics to perform inspection from outside of the confined space. However, teleoperation of the crane robot can still be a challenge due to the need to avoid obstacles in the workspace and potential oscillations of the camera payload. The main contribution of this work is to exploit the differential flatness of the crane-robot dynamics for designing reduced-oscillation, collision-free time trajectories of the camera payload for use in teleoperation. Autonomous experiments verify the efficacy of removing undesired oscillations by 89%. Furthermore, teleoperation experiments demonstrate that the controller eliminated collisions (from 33% to 0%) when 12 participants performed an inspection task with the use of proposed trajectory selection when compared to the case without it. Moreover, even discounting the failures due to collisions, the proposed approach improved task efficiency by 18.7% when compared to the case without it.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2742-2749"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143379604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This letter presents a novel weakly supervised semantic segmentation method for radar segmentation, where the existing LiDAR semantic segmentation models are employed to generate semantic labels, which then serve as supervision signals for training a radar semantic segmentation model. The obtained radar semantic segmentation model maintains consistent and robust segmentation performance under all-weather conditions, particularly in the snow, rain and fog. To mitigate potential errors in LiDAR semantic labels, we design a dedicated refinement scheme that corrects erroneous labels based on structural features and distribution patterns. The semantic information generated by our radar segmentation model is used in two downstream tasks, achieving significant performance improvements. In large-scale radar-based localization using OpenStreetMap, it leads to localization error reduction by 20.55% over prior methods. For the odometry task, it improves translation accuracy by 16.4% compared to the second-best method, securing the first place in the radar odometry competition at the Radar in Robotics workshop of ICRA 2024, Japan.
{"title":"Get It for Free: Radar Segmentation Without Expert Labels and Its Application in Odometry and Localization","authors":"Siru Li;Ziyang Hong;Yushuai Chen;Liang Hu;Jiahu Qin","doi":"10.1109/LRA.2025.3536196","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536196","url":null,"abstract":"This letter presents a novel weakly supervised semantic segmentation method for radar segmentation, where the existing LiDAR semantic segmentation models are employed to generate semantic labels, which then serve as supervision signals for training a radar semantic segmentation model. The obtained radar semantic segmentation model maintains consistent and robust segmentation performance under all-weather conditions, particularly in the snow, rain and fog. To mitigate potential errors in LiDAR semantic labels, we design a dedicated refinement scheme that corrects erroneous labels based on structural features and distribution patterns. The semantic information generated by our radar segmentation model is used in two downstream tasks, achieving significant performance improvements. In large-scale radar-based localization using OpenStreetMap, it leads to localization error reduction by 20.55% over prior methods. For the odometry task, it improves translation accuracy by 16.4% compared to the second-best method, securing the first place in the radar odometry competition at the Radar in Robotics workshop of ICRA 2024, Japan.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2678-2685"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143361297","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3534681
Woojae Lee;Taehyun Kim;Jeongeun Kim;TaeWon Seo
Differential-driven wheeled mobile robots, such as logistics robots and mobile manipulators, are used for various tasks on flat ground. These differential-driven wheeled mobile robots are highly environmentally constrained when driven on flat ground. This study proposes a differential-driven wheeled mobile mechanism of the robot with high step-climbing ability and the capability to navigate narrow paths. This work shows that the novel wheel and differential-driven module improve the ability to climb stairs through the transfer of center of mass (C.O.M). A sub-wheel connected to the passive joint of the wheel is used to convert the drive into a vertical force to improve the mobile robot's ability to climb high stairs. By pitching the body of the wheel-powered robot through the advanced reaction force, the ability to climb stairs in reverse using the reaction force of the wall is improved. Furthermore, the proposed mechanism enables precise control of the robot's path, allowing it to effectively navigate narrow spaces. The prototype robot was tested for climbing stairs and high steps, as well as climbing a deformable slope while climbing obstacles. Even if the center of mass is located in the driving direction, this result uses a novel wheel to overcome the step in the front wheel and improve the overcoming performance of the rear wheel owing to the center-of-mass position in the driving direction. We expect that this method will be applicable to various differential-driven wheeled mobile robot mechanisms, especially for environments with confined spaces.
{"title":"Differential-Driven Wheeled Mobile Robot Mechanism With High Step-Climbing Ability","authors":"Woojae Lee;Taehyun Kim;Jeongeun Kim;TaeWon Seo","doi":"10.1109/LRA.2025.3534681","DOIUrl":"https://doi.org/10.1109/LRA.2025.3534681","url":null,"abstract":"Differential-driven wheeled mobile robots, such as logistics robots and mobile manipulators, are used for various tasks on flat ground. These differential-driven wheeled mobile robots are highly environmentally constrained when driven on flat ground. This study proposes a differential-driven wheeled mobile mechanism of the robot with high step-climbing ability and the capability to navigate narrow paths. This work shows that the novel wheel and differential-driven module improve the ability to climb stairs through the transfer of center of mass (C.O.M). A sub-wheel connected to the passive joint of the wheel is used to convert the drive into a vertical force to improve the mobile robot's ability to climb high stairs. By pitching the body of the wheel-powered robot through the advanced reaction force, the ability to climb stairs in reverse using the reaction force of the wall is improved. Furthermore, the proposed mechanism enables precise control of the robot's path, allowing it to effectively navigate narrow spaces. The prototype robot was tested for climbing stairs and high steps, as well as climbing a deformable slope while climbing obstacles. Even if the center of mass is located in the driving direction, this result uses a novel wheel to overcome the step in the front wheel and improve the overcoming performance of the rear wheel owing to the center-of-mass position in the driving direction. We expect that this method will be applicable to various differential-driven wheeled mobile robot mechanisms, especially for environments with confined spaces.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2702-2709"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143361302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3534523
Ayca Takmaz;Alexandros Delitzas;Robert W. Sumner;Francis Engelmann;Johanna Wald;Federico Tombari
Open-vocabulary 3D segmentation enables exploration of 3D spaces using free-form text descriptions. Existing methods for open-vocabulary 3D instance segmentation primarily focus on identifying object-level instances but struggle with finer-grained scene entities such as object parts, or regions described by generic attributes. In this work, we introduce Search3D, an approach to construct hierarchical open-vocabulary 3D scene representations, enabling 3D search at multiple levels of granularity: fine-grained object parts, entire objects, or regions described by attributes like materials. Unlike prior methods, Search3D shifts towards a more flexible open-vocabulary 3D search paradigm, moving beyond explicit object-centric queries. For systematic evaluation, we further contribute a scene-scale open-vocabulary 3D part segmentation benchmark based on MultiScan, along with a set of open-vocabulary fine-grained part annotations on ScanNet++. Search3D outperforms baselines in scene-scale open-vocabulary 3D part segmentation, while maintaining strong performance in segmenting 3D objects and materials.
{"title":"Search3D: Hierarchical Open-Vocabulary 3D Segmentation","authors":"Ayca Takmaz;Alexandros Delitzas;Robert W. Sumner;Francis Engelmann;Johanna Wald;Federico Tombari","doi":"10.1109/LRA.2025.3534523","DOIUrl":"https://doi.org/10.1109/LRA.2025.3534523","url":null,"abstract":"Open-vocabulary 3D segmentation enables exploration of 3D spaces using free-form text descriptions. Existing methods for open-vocabulary 3D instance segmentation primarily focus on identifying <italic>object</i>-level instances but struggle with finer-grained scene entities such as <italic>object parts</i>, or regions described by generic <italic>attributes</i>. In this work, we introduce Search3D, an approach to construct hierarchical open-vocabulary 3D scene representations, enabling 3D search at multiple levels of granularity: fine-grained object parts, entire objects, or regions described by attributes like materials. Unlike prior methods, Search3D shifts towards a more flexible open-vocabulary 3D search paradigm, moving beyond explicit object-centric queries. For systematic evaluation, we further contribute a scene-scale open-vocabulary 3D part segmentation benchmark based on MultiScan, along with a set of open-vocabulary fine-grained part annotations on ScanNet++. Search3D outperforms baselines in scene-scale open-vocabulary 3D part segmentation, while maintaining strong performance in segmenting 3D objects and materials.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2558-2565"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143361465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3536221
Ameer Helmi;Tze-Hsuan Wang;Samuel W. Logan;Naomi T. Fitter
Young children with motor disabilities face barriers and delays to learning motor skills such as walking. Pediatric body-weight support harness systems (BWSHes) are a newer technology for helping young children to practice supported motor skills. Incorporating an assistive robot to mediate BWSH interventions can support further child motion and engagement, but almost no work to date has studied autonomous robot-mediated BWSH use. We conducted a six-month-long single-case study series with two participants to evaluate the effectiveness of an autonomous assistive robot in motivating the children to move and stay engaged while in the BWSH. We collected and analyzed objective movement data and self-reported parent survey data to determine how much the child moved and stayed engaged during sessions. Our results showed that both children displayed more movement while the assistive robot was active (relative to in prior no-robot periods). Parents also rated their children as more engaged while the assistive robot was present. An autonomous assistive robot may provide motivation for a child to move and stay engaged while using a pediatric rehabilitation aid such as a BWSH. The products of this work can benefit roboticists who work with children with disabilities and researchers who use pediatric rehabilitation technologies.
{"title":"Look at Them Go! Using an Autonomous Assistive GoBot to Encourage Movement Practice by Two Children With Motor Disabilities","authors":"Ameer Helmi;Tze-Hsuan Wang;Samuel W. Logan;Naomi T. Fitter","doi":"10.1109/LRA.2025.3536221","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536221","url":null,"abstract":"Young children with motor disabilities face barriers and delays to learning motor skills such as walking. Pediatric body-weight support harness systems (BWSHes) are a newer technology for helping young children to practice supported motor skills. Incorporating an assistive robot to mediate BWSH interventions can support further child motion and engagement, but almost no work to date has studied autonomous robot-mediated BWSH use. We conducted a six-month-long single-case study series with two participants to evaluate the effectiveness of an autonomous assistive robot in motivating the children to move and stay engaged while in the BWSH. We collected and analyzed objective movement data and self-reported parent survey data to determine how much the child moved and stayed engaged during sessions. Our results showed that both children displayed more movement while the assistive robot was active (relative to in prior no-robot periods). Parents also rated their children as more engaged while the assistive robot was present. An autonomous assistive robot may provide motivation for a child to move and stay engaged while using a pediatric rehabilitation aid such as a BWSH. The products of this work can benefit roboticists who work with children with disabilities and researchers who use pediatric rehabilitation technologies.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 4","pages":"3318-3325"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Robust depth perception in visually-degraded environments is crucial for autonomous aerial systems. Thermal imaging cameras, which capture infrared radiation, are robust to visual degradation. However, due to lack of a large-scale dataset, the use of thermal cameras for uncrewed aerial system (UAS) depth perception has remained largely unexplored. This paper presents a stereo thermal depth perception dataset for autonomous aerial perception applications. The dataset consists of stereo thermal images, LiDAR, IMU and ground truth depth maps captured in urban and forest settings under diverse conditions like day, night, rain, and smoke. We benchmark representative stereo depth estimation algorithms, offering insights into their performance in degraded conditions. Models trained on our dataset generalize well to unseen smoky conditions, highlighting the robustness of stereo thermal imaging for depth perception. We aim for this work to enhance robotic perception in disaster scenarios, allowing for exploration and operations in previously unreachable areas.
{"title":"FIReStereo: Forest InfraRed Stereo Dataset for UAS Depth Perception in Visually Degraded Environments","authors":"Devansh Dhrafani;Yifei Liu;Andrew Jong;Ukcheol Shin;Yao He;Tyler Harp;Yaoyu Hu;Jean Oh;Sebastian Scherer","doi":"10.1109/LRA.2025.3536278","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536278","url":null,"abstract":"Robust depth perception in visually-degraded environments is crucial for autonomous aerial systems. Thermal imaging cameras, which capture infrared radiation, are robust to visual degradation. However, due to lack of a large-scale dataset, the use of thermal cameras for uncrewed aerial system (UAS) depth perception has remained largely unexplored. This paper presents a stereo thermal depth perception dataset for autonomous aerial perception applications. The dataset consists of stereo thermal images, LiDAR, IMU and ground truth depth maps captured in urban and forest settings under diverse conditions like day, night, rain, and smoke. We benchmark representative stereo depth estimation algorithms, offering insights into their performance in degraded conditions. Models trained on our dataset generalize well to unseen smoky conditions, highlighting the robustness of stereo thermal imaging for depth perception. We aim for this work to enhance robotic perception in disaster scenarios, allowing for exploration and operations in previously unreachable areas.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 4","pages":"3302-3309"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489205","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1109/LRA.2025.3536294
Seungwon Choi;Tae-Wan Kim
Robust state estimation is a fundamental research topic in robotics. Existing approaches like robust kernels combined with iteratively re-weighted least squares (IRLS) often require heuristic parameter selection and extensive fine-tuning. In this manuscript, we propose a novel method that optimizes kernels while preserving the advantages of existing techniques. By introducing a probabilistic interpretation of weights and residuals, our approach enables automatic parameter selection. Applied to iterative closest point (ICP) and bundle adjustment (BA), experimental results demonstrate improved convergence and robustness compared to traditional methods, eliminating the need for time-consuming parameter tuning and offering a practical solution for robust state estimation.
{"title":"Probabilistic Kernel Optimization for Robust State Estimation","authors":"Seungwon Choi;Tae-Wan Kim","doi":"10.1109/LRA.2025.3536294","DOIUrl":"https://doi.org/10.1109/LRA.2025.3536294","url":null,"abstract":"Robust state estimation is a fundamental research topic in robotics. Existing approaches like robust kernels combined with iteratively re-weighted least squares (IRLS) often require heuristic parameter selection and extensive fine-tuning. In this manuscript, we propose a novel method that optimizes kernels while preserving the advantages of existing techniques. By introducing a probabilistic interpretation of weights and residuals, our approach enables automatic parameter selection. Applied to iterative closest point (ICP) and bundle adjustment (BA), experimental results demonstrate improved convergence and robustness compared to traditional methods, eliminating the need for time-consuming parameter tuning and offering a practical solution for robust state estimation.","PeriodicalId":13241,"journal":{"name":"IEEE Robotics and Automation Letters","volume":"10 3","pages":"2998-3005"},"PeriodicalIF":4.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10857458","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143430574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：