Perception Stitching enables zero-shot transfer of visuomotor policies, significantly improving success rates.

Pingcheng Jian, Easop Lee, Zachary Bell et al.

Dreamitate fine-tunes a video diffusion model for robot visuomotor tasks, outperforming behavior cloning in generalization, with success rates over 85% in four tasks.

Junbang Liang, Ruoshi Liu, Ege Ozguroglu et al.

OmniH2O employs goal-conditioned imitation learning and multimodal interfaces to enable dexterous whole-body humanoid teleoperation and autonomous control.

Tairan He, Zhengyi Luo, Xialin He et al.

This study systematically evaluates multiple high-rated LLMs in robotics, revealing significant biases and safety risks, including discrimination and acceptance of unlawful commands.

Andrew Hundt, Rumaisa Azeem, Masoumeh Mansouri et al.

RVT-2 employs multi-stage virtual view reasoning and system-level optimizations to achieve 82% success in high-precision multi-task manipulation with only 10 demonstrations, training 6× faster than RVT.

Ankit Goyal, Valts Blukis, Jie Xu et al.

BAKU combines multimodal conditioning and action chunking, reaching 90% on LIBERO-90 and 91% on real xArm tasks.

Siddhant Haldar, Zhuoran Peng, Lerrel Pinto

Proposed CoBL-Diffusion integrates Control Barrier and Lyapunov functions into diffusion models for safe robot planning in dynamic environments.

Kazuki Mizuta, Karen Leung

SDP accelerates policy synthesis by partial denoising of action trajectories, reducing sampling time by 25%.

Sigmund H. Høeg, Yilun Du, Olav Egeland

Bench2Drive is a multi-ability closed-loop autonomous driving benchmark with 2 million annotated frames, evaluating 44 scenarios in CARLA.

Xiaosong Jia, Zhenjie Yang, Qifeng Li et al.

POAM combines non-stationary attentive kernels with variational EM for constant-time online mapping in robotic exploration.

Weizhe Chen, Lantao Liu, Roni Khardon

RoboCasa leverages generative AI and large-scale simulation for multi-task robot training in household environments, achieving 47.6% success in complex tasks.

Soroush Nasiriany, Abhiram Maddukuri, Lance Zhang et al.

Proposes Video-Language Critic, a contrastive video-text model, achieving 2x sample efficiency in Meta-World tasks and cross-embodiment transfer.

Minttu Alakuijala, Reginald McLean, Isaac Woungang et al.

InterPreT leverages GPT-4 to learn symbolic predicates from language feedback, enabling robots to generalize long-horizon planning with minimal manual design.

Muzhi Han, Yifeng Zhu, Song-Chun Zhu et al.

Proposes FRPN guidance and IMM filtering for fast, accurate UAV mid-air interception, reducing response time by 20% and increasing success rate to 85%.

Michal Pliska, Matouš Vrba, Tomáš Báča et al.

Proposes a resilient planning model based on FA-MDP, incorporating actuator failure probabilities, to optimize robot fault response strategies.

Kyle Baldes, Diptanil Chaudhuri, Jason M. O'Kane et al.

Multi-Q learning with dual agents optimizes UAV path and connectivity in challenging terrains, achieving over 90% success rate with low communication outages.

Mohammed M. H. Qazzaz, Syed A. R. Zaidi, Desmond C. McLernon et al.

COAST is a sampling-based algorithm combining stream motion planning with constrained task planning, achieving an order-of-magnitude speedup in complex robotic tasks.

Brandon Vu, Toki Migimatsu, Jeannette Bohg

Proposed Consistency Policy accelerates visuomotor control by 10x via self-consistency distillation, maintaining high success rates.

Aaditya Prasad, Kevin Lin, Jimmy Wu et al.

RoboHop introduces a segment-based topological map with semantic descriptors, enabling open-vocabulary queries and zero-shot navigation in complex environments.

Sourav Garg, Krishan Rana, Mehdi Hosseinzadeh et al.

Proposes a multimodal deep learning framework for robust place recognition, achieving 85% Top-1 recall on KITTI dataset.

Peng Yin, Jianhao Jiao, Shiqi Zhao et al.