robotic-manipulation

Rolling-WAM introduces a method to distribute denoising across replanning cycles in world action models for robotic manipulation, achieving a 4.5x speedup in replanning while maintaining competitive performance.

Introduces Movement Trend Guidance to enhance 3D diffusion policies in robotic manipulation by providing foresight without explicit trajectories, achieving improved performance on benchmarks like RoboTwin2.0 and LIBERO-40.

A general-purpose agent directly controls a physical robot by interpreting visuals, writing executable programs, and revising actions based on physical feedback across diverse manipulation tasks, achieving high success rates without task-specific training.

MaP-WAM introduces a framework for non-Markovian robotic manipulation by separating memory-grounded planning from execution, using episodic memory to maintain fixed inference latency and achieving state-of-the-art results.

Researchers enable simple robotic grippers to achieve precise in-hand object manipulation through controlled sliding using tactile sensing and advanced friction models, mimicking human dexterity.

This paper introduces GE-Act 2.0, a world-action model pretrained from scratch to enable scalable zero-shot robotic manipulation with improved success rates across diverse tasks and conditions.

ZimaBlue introduces a scalable framework for learning generalizable world action models from large-scale egocentric video, substantially improving zero-shot robotic manipulation through a three-stage curriculum and slow-fast architecture.

PhysCaP is a physics-informed code-generation agent that actively explores objects to infer hidden physical properties for efficient robotic manipulation.

GOAG is a deep generative grasp planner that learns a gripper-specific contact surface distribution to sample valid grasps for unseen objects without object-specific training, achieving state-of-the-art results on the MultiDex dataset for dexterous robotic manipulation.

DreamX-Phi 1.0 is an action-conditioned video world model for robotic manipulation that predicts future observations using geometric attention encoding, depth estimation, and distillation, achieving top results in the WorldArena 2.0 Challenge.

TurboVLA introduces a new Vision-Language-Action paradigm that directly maps vision and language to action, achieving 97.7% success on LIBERO with only 0.2B parameters and real-time inference at 32 Hz on consumer GPUs, significantly reducing computational cost.

TableVerse introduces a fully automated Real2Sim pipeline that converts unstructured, in-the-wild images into high-fidelity, simulation-ready tabletop environments with accurate metrics and physical stability, along with a large-scale dataset (TableVerse-100K) for generalizable robotic manipulation.

LaMem-VLA proposes a latent-memory-native framework that integrates short-term and long-term historical experience directly into Vision-Language-Action reasoning, enabling better performance on long-horizon robotic manipulation tasks.

RynnWorld-4D is a generative world model that co-produces future RGB, depth, and optical flow from a single RGB-D image and language instruction using a unified diffusion process, enabling efficient robotic manipulation through inverse dynamics policy learning. It achieves state-of-the-art on real-world bimanual manipulation tasks.

PhysisForcing is a training framework that enhances embodied video generation for robotic manipulation by enforcing physical consistency through pixel-level trajectory alignment and semantic-level relational alignment losses in a DiT-based architecture, achieving notable improvements on benchmarks.

Foresight is a failure detection framework for long-horizon robotic manipulation that uses action-conditioned world model latents and functional conformal prediction to monitor trajectories, trained only with final task labels. It demonstrates state-of-the-art performance across simulation and real robot tasks.

EventVLA introduces a sparse visual evidence memory framework for long-horizon robotic manipulation, achieving an average success rate improvement of +40% over state-of-the-art memory-augmented VLAs.

Presents Qwen-RobotManip, a Vision-Language-Action foundation model for robotic manipulation that achieves generalization through unified alignment across representation, motion, and behavior dimensions, enabling large-scale training on diverse data sources. It outperforms prior state-of-the-art models across multiple out-of-distribution benchmarks and demonstrates emergent capabilities like zero-shot instruction following and cross-embodiment transfer.

PAIWorld enhances diffusion-transformer world models with geometric awareness and cross-view attention to improve multi-view 3D consistency for robotic manipulation tasks, achieving state-of-the-art results on benchmarks.

WEAVER is a multi-view world model for robotic manipulation that achieves high fidelity, consistency, and efficiency using flow-matching loss, demonstrating superior performance in policy evaluation, improvement, and test-time planning with significant real-world improvements.