Proposes a variational multimodal representation learning framework integrating RGB-D, force, and proprioception, boosting contact-rich task sample efficiency by 20%.

Michelle A. Lee, Yuke Zhu, Peter Zachares et al.

Applying normalized DTW (nDTW) for path similarity evaluation improves human correlation and RL training in navigation tasks.

Gabriel Ilharco, Vihan Jain, Alexander Ku et al.

Proposes a particle filter-based active perception framework for robot model uncertainty estimation, improving manipulation accuracy.

Oliver Kroemer, Scott Niekum, George Konidaris

Combines heuristic A* search and B-spline optimization for fast, safe quadrotor trajectory generation in complex environments.

Boyu Zhou, Fei Gao, Luqi Wang et al.

Introducing VICES as an action space for deep RL, enhancing sample efficiency, safety, and transferability in contact-rich tasks.

Roberto Martín-Martín, Michelle A. Lee, Rachel Gardner et al.

Combines Bayesian Eigenobjects with language models for viewpoint-invariant object understanding, enabling few-shot learning and attribute prediction.

Vanya Cohen, Benjamin Burchfiel, Thao Nguyen et al.

FASTER achieves safe flights up to 3.6 m/s in unknown environments.

Jesus Tordesillas, Brett T. Lopez, Jonathan P. How

Proposes Play-LMP, a self-supervised latent plan model from human play data, achieving superior generalization in multi-task robotic manipulation with 85.5% success rate.

Corey Lynch, Mohi Khansari, Ted Xiao et al.

Convex optimization-based dynamic parameter identification for dVRK, ensuring physical consistency and improved accuracy.

Yan Wang, Radian Gondokaryono, Adnan Munawar et al.

Proposes OA-IPP, a Gaussian Process-based adaptive path planning algorithm for UAV target search in cluttered environments, outperforming state-of-the-art methods.

Ajith Anil Meera, Marija Popovic, Alexander Millane et al.

Combining YOLOv2-based detection with point cloud clustering, the system achieves real-time 3D pointing recognition with <10° error at 30Hz on standard hardware.

Bita Azari, Angelica Lim, Richard T. Vaughan

Simulation-trained deep neural policies enable agile, energy-efficient locomotion and fall recovery in a large quadruped robot, with direct sim-to-real transfer.

Jemin Hwangbo, Joonho Lee, Alexey Dosovitskiy et al.

MID-Fusion employs octree-based object-level volumetric mapping for multi-instance dynamic SLAM, integrating geometric, semantic, and motion data.

Binbin Xu, Wenbin Li, Dimos Tzoumanikas et al.

Residual RL combines classical feedback control with deep RL to improve robustness and efficiency in contact-rich robot tasks, reducing training time and increasing success rates.

Tobias Johannink, Shikhar Bahl, Ashvin Nair et al.

Introduces a micro Lie theory for state estimation in robotics, simplifying complexities in SLAM and visual odometry.

Joan Solà, Jeremie Deray, Dinesh Atchuthan

Proposes a graph-based redundant robot assignment framework using supermodular optimization to reduce waiting times under uncertain path costs.

Amanda Prorok

Large-scale comparison of 14 robust filters for ICP point cloud registration reveals stability and parameter effects across diverse environments.

Philippe Babin, Philippe Giguère, François Pomerleau

Proposes a lightweight snake gate detection algorithm combined with improved pose estimation and predictive control, enabling autonomous drone racing at 1.5m/s with minimal error.

S. Li, M. M. O. I. Ozo, C. De Wagter et al.

Baidu Apollo EM motion planner combines multi-lane path and speed optimization using DP and spline QP, ensuring safety and comfort.

Haoyang Fan, Fan Zhu, Changchun Liu et al.

Proposed TOG-Net, a framework for task-oriented grasping via simulated self-supervision; achieved 71.1% sweeping and 80.0% hammering success rates.

Kuan Fang, Yuke Zhu, Animesh Garg et al.