JOPAT jointly models pixels, point tracks, and actions in a diffusion transformer and reports gains over pixel-only baselines on long-horizon robot tasks with occlusion and off-screen motion.
hub Canonical reference
ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation
Canonical reference. 71% of citing Pith papers cite this work as background.
27
Pith papers citing it
Background
71% of classified citations
abstract
Representing robotic manipulation tasks as constraints that associate the robot and the environment is a promising way to encode desired robot behaviors. However, it remains unclear how to formulate the constraints such that they are 1) versatile to diverse tasks, 2) free of manual labeling, and 3) optimizable by off-the-shelf solvers to produce robot actions in real-time. In this work, we introduce Relational Keypoint Constraints (ReKep), a visually-grounded representation for constraints in robotic manipulation. Specifically, ReKep is expressed as Python functions mapping a set of 3D keypoints in the environment to a numerical cost. We demonstrate that by representing a manipulation task as a sequence of Relational Keypoint Constraints, we can employ a hierarchical optimization procedure to solve for robot actions (represented by a sequence of end-effector poses in SE(3)) with a perception-action loop at a real-time frequency. Furthermore, in order to circumvent the need for manual specification of ReKep for each new task, we devise an automated procedure that leverages large vision models and vision-language models to produce ReKep from free-form language instructions and RGB-D observations. We present system implementations on a wheeled single-arm platform and a stationary dual-arm platform that can perform a large variety of manipulation tasks, featuring multi-stage, in-the-wild, bimanual, and reactive behaviors, all without task-specific data or environment models. Website at https://rekep-robot.github.io/.
hub tools
citation-role summary
background 11
baseline 1
method 1
other 1
citation-polarity summary
representative citing papers
CreFlow combines LTL compositional rewards with credit-aware NFT and corrective reflow losses in online RL to improve embodied video diffusion models, raising downstream task success by 23.8 percentage points on eight bimanual manipulation tasks.
PaMoSplat reconstructs dynamic scenes by lifting 2D segmentations to coherent 3D Gaussian parts and estimating their motions via optical flow-guided differential evolution for higher quality rendering and faster training.
KITE is a training-free method that uses keyframe-indexed tokenized evidence including BEV schematics to enhance VLM performance on robot failure detection, identification, localization, explanation, and correction.
ST-BiBench reveals a coordination paradox in which MLLMs show strong high-level strategic reasoning yet fail at fine-grained 16-dimensional bimanual action synthesis and multi-stream fusion.
AVP architecture has VLM emit visual-primitive tokens to condition flow-matching action expert, yielding 27.61% higher success rate than pi_0.5 on real-robot pick-and-place tasks.
AgentChord models manipulation tasks as directed graphs enriched with anticipatory recovery branches, using specialized agents to enable immediate, low-latency failure responses and improve success on long-horizon bimanual tasks.
TriRelVLA introduces triadic object-hand-task relational representations and a task-grounded graph transformer with a relational bottleneck to improve generalization in robotic manipulation across scenes, objects, and tasks.
Decompose and Recompose decomposes seen robotic demonstrations into skill-action alignments and recomposes them via visual-semantic retrieval and planning to enable zero-shot cross-task generalization.
BridgeACT learns robot manipulation from human videos alone by predicting task-relevant grasp regions and 3D motion affordances that map directly to robot controllers.
CorridorVLA improves VLA models by using predicted sparse anchors to impose explicit spatial corridors on action trajectories, yielding 3.4-12.4% success rate gains on LIBERO-Plus with GR00T-Corr reaching 83.21%.
PALM improves long-horizon robotic manipulation success by distilling affordance representations for object interaction and predicting within-subtask progress in a VLA model.
IGen generates realistic visuomotor training data including actions and temporally coherent visuals from unstructured open-world images via 3D reconstruction and VLM reasoning.
InternVLA-M1 uses spatially guided pre-training on 2.3M examples followed by action post-training to deliver up to 17% gains on robot manipulation benchmarks and 20.6% on unseen objects.
COMPASS is a manipulation-aware active sensing framework that raises simulated manipulation success rates by 24.25% over information-gain-only baselines in a new four-level confined-space benchmark.
Embodied-R1 uses a pointing-centric representation and reinforced fine-tuning on a 200K dataset to achieve state-of-the-art results on embodied benchmarks plus 56.2% success in SIMPLEREnv and 87.5% on real XArm tasks without task-specific training.
RIGVid shows that filtered AI-generated videos can serve as effective supervision for complex robotic manipulation tasks without any real demonstrations.
CoT-VLA is a 7B VLA that generates future visual frames autoregressively as planning goals before actions, outperforming prior VLAs by 17% on real-world tasks and 6% in simulation.
HybridVLA unifies diffusion and autoregression in a single VLA model via collaborative training and ensemble to raise robot manipulation success rates by 14% in simulation and 19% in real-world tasks.
FAST applies discrete cosine transform to robot action sequences for efficient tokenization, enabling autoregressive VLAs to succeed on high-frequency dexterous tasks and scale to 10k hours of data while matching diffusion VLA performance with up to 5x faster training.
This is the first survey on vision-language-action models, providing a taxonomy across three lines, plus summaries of datasets, simulators, benchmarks, challenges, and future directions in embodied AI.
Forecast-GS predicts task-completed 3D states via Gaussian splatting to achieve higher success rates than baselines in real-world language-conditioned manipulation tasks.
BioProVLA-Agent integrates protocol parsing, visual state verification, and VLA-based execution in a closed-loop multi-agent framework with AugSmolVLA augmentation to improve robustness for biological lab tasks like tube handling and liquid pouring.
A framework integrates anticipatory planning and real-time feedback via reliability-aware optimization and phase switching to achieve efficient, reliable continuous mobile manipulation under uncertainty.
citing papers explorer
-
Point Tracking Improves World Action Models
JOPAT jointly models pixels, point tracks, and actions in a diffusion transformer and reports gains over pixel-only baselines on long-horizon robot tasks with occlusion and off-screen motion.
-
CreFlow: Corrective Reflow for Sparse-Reward Embodied Video Diffusion RL
CreFlow combines LTL compositional rewards with credit-aware NFT and corrective reflow losses in online RL to improve embodied video diffusion models, raising downstream task success by 23.8 percentage points on eight bimanual manipulation tasks.
-
PaMoSplat: Part-Aware Motion-Guided Gaussian Splatting for Dynamic Scene Reconstruction
PaMoSplat reconstructs dynamic scenes by lifting 2D segmentations to coherent 3D Gaussian parts and estimating their motions via optical flow-guided differential evolution for higher quality rendering and faster training.
-
KITE: Keyframe-Indexed Tokenized Evidence for VLM-Based Robot Failure Analysis
KITE is a training-free method that uses keyframe-indexed tokenized evidence including BEV schematics to enhance VLM performance on robot failure detection, identification, localization, explanation, and correction.
-
ST-BiBench: Benchmarking Multi-Stream Multimodal Coordination in Bimanual Embodied Tasks for MLLMs
ST-BiBench reveals a coordination paradox in which MLLMs show strong high-level strategic reasoning yet fail at fine-grained 16-dimensional bimanual action synthesis and multi-stream fusion.
-
Action with Visual Primitives
AVP architecture has VLM emit visual-primitive tokens to condition flow-matching action expert, yielding 27.61% higher success rate than pi_0.5 on real-robot pick-and-place tasks.
-
From Reaction to Anticipation: Proactive Failure Recovery through Agentic Task Graph for Robotic Manipulation
AgentChord models manipulation tasks as directed graphs enriched with anticipatory recovery branches, using specialized agents to enable immediate, low-latency failure responses and improve success on long-horizon bimanual tasks.
-
TriRelVLA: Triadic Relational Structure for Generalizable Embodied Manipulation
TriRelVLA introduces triadic object-hand-task relational representations and a task-grounded graph transformer with a relational bottleneck to improve generalization in robotic manipulation across scenes, objects, and tasks.
-
Decompose and Recompose: Reasoning New Skills from Existing Abilities for Cross-Task Robotic Manipulation
Decompose and Recompose decomposes seen robotic demonstrations into skill-action alignments and recomposes them via visual-semantic retrieval and planning to enable zero-shot cross-task generalization.
-
BridgeACT: Bridging Human Demonstrations to Robot Actions via Unified Tool-Target Affordances
BridgeACT learns robot manipulation from human videos alone by predicting task-relevant grasp regions and 3D motion affordances that map directly to robot controllers.
-
CorridorVLA: Explicit Spatial Constraints for Generative Action Heads via Sparse Anchors
CorridorVLA improves VLA models by using predicted sparse anchors to impose explicit spatial corridors on action trajectories, yielding 3.4-12.4% success rate gains on LIBERO-Plus with GR00T-Corr reaching 83.21%.
-
PALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic Manipulation
PALM improves long-horizon robotic manipulation success by distilling affordance representations for object interaction and predicting within-subtask progress in a VLA model.
-
IGen: Scalable Data Generation for Robot Learning from Open-World Images
IGen generates realistic visuomotor training data including actions and temporally coherent visuals from unstructured open-world images via 3D reconstruction and VLM reasoning.
-
InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy
InternVLA-M1 uses spatially guided pre-training on 2.3M examples followed by action post-training to deliver up to 17% gains on robot manipulation benchmarks and 20.6% on unseen objects.
-
COMPASS: Confined-space Manipulation Planning with Active Sensing Strategy
COMPASS is a manipulation-aware active sensing framework that raises simulated manipulation success rates by 24.25% over information-gain-only baselines in a new four-level confined-space benchmark.
-
Embodied-R1: Reinforced Embodied Reasoning for General Robotic Manipulation
Embodied-R1 uses a pointing-centric representation and reinforced fine-tuning on a 200K dataset to achieve state-of-the-art results on embodied benchmarks plus 56.2% success in SIMPLEREnv and 87.5% on real XArm tasks without task-specific training.
-
Robotic Manipulation by Imitating Generated Videos Without Physical Demonstrations
RIGVid shows that filtered AI-generated videos can serve as effective supervision for complex robotic manipulation tasks without any real demonstrations.
-
CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models
CoT-VLA is a 7B VLA that generates future visual frames autoregressively as planning goals before actions, outperforming prior VLAs by 17% on real-world tasks and 6% in simulation.
-
HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model
HybridVLA unifies diffusion and autoregression in a single VLA model via collaborative training and ensemble to raise robot manipulation success rates by 14% in simulation and 19% in real-world tasks.
-
FAST: Efficient Action Tokenization for Vision-Language-Action Models
FAST applies discrete cosine transform to robot action sequences for efficient tokenization, enabling autoregressive VLAs to succeed on high-frequency dexterous tasks and scale to 10k hours of data while matching diffusion VLA performance with up to 5x faster training.
-
A Survey on Vision-Language-Action Models for Embodied AI
This is the first survey on vision-language-action models, providing a taxonomy across three lines, plus summaries of datasets, simulators, benchmarks, challenges, and future directions in embodied AI.
-
Forecast-aware Gaussian Splatting for Predictive 3D Representation in Language-Guided Pick-and-Place Manipulation
Forecast-GS predicts task-completed 3D states via Gaussian splatting to achieve higher success rates than baselines in real-world language-conditioned manipulation tasks.
-
BioProVLA-Agent: An Affordable, Protocol-Driven, Vision-Enhanced VLA-Enabled Embodied Multi-Agent System with Closed-Loop-Capable Reasoning for Biological Laboratory Manipulation
BioProVLA-Agent integrates protocol parsing, visual state verification, and VLA-based execution in a closed-loop multi-agent framework with AugSmolVLA augmentation to improve robustness for biological lab tasks like tube handling and liquid pouring.
-
Synergizing Efficiency and Reliability for Continuous Mobile Manipulation
A framework integrates anticipatory planning and real-time feedback via reliability-aware optimization and phase switching to achieve efficient, reliable continuous mobile manipulation under uncertainty.
-
FUNCanon: Learning Pose-Aware Action Primitives via Functional Object Canonicalization for Generalizable Robotic Manipulation
FunCanon introduces functional object canonicalization with VLM affordances to create pose-aware action primitives for generalizable imitation learning in robotic manipulation.
- AssemLM: A Spatial Reasoning Multimodal Large Language Model for Robotic Assembly
- From Video to Control: A Survey of Learning Manipulation Interfaces from Temporal Visual Data