archive
Every paper Pith has read. Search by title, abstract, or pith.
2900 papers in cs.RO · page 16
-
Mixture planning and ensemble returns stabilize long-horizon visual control
ELVIS: Ensemble-Calibrated Latent Imagination for Long-Horizon Visual MPC
-
Latent actions in VLAs match type to task for gains
From Pixels to Tokens: A Systematic Study of Latent Action Supervision for Vision-Language-Action Models
-
AI learning refines dead-reckoning for accurate AUV paths
AI-Aided Advancements in Autonomous Underwater Vehicle Navigation
-
Tactile RL hits 67% success at 0.05 mm peg insertions
From Reach to Insert: Tactile-Augmented Precision Assembly under Sub-Millimeter Tolerances
-
RL fine-tuning unlocks self-editing in discrete diffusion driving
ReflectDrive-2: Reinforcement-Learning-Aligned Self-Editing for Discrete Diffusion Driving
-
RL training lets driving diffusion model edit its own trajectories
ReflectDrive-2: Reinforcement-Learning-Aligned Self-Editing for Discrete Diffusion Driving
-
Active motions let robots confirm firm grasps before handover
Active Contact Sensing for Robust Robot-to-Human Object Handover
-
Hybrid-fidelity MPC runs bipedal walking in real time
Right Model, Right Time: Real-Time Cascaded-Fidelity MPC for Bipedal Walking
-
Gradient ascent on policy trajectories beats gradient-free MPC
Dream-MPC: Gradient-Based Model Predictive Control with Latent Imagination
-
Gradient ascent on policy trajectories lifts MPC performance in RL
Dream-MPC: Gradient-Based Model Predictive Control with Latent Imagination
-
Binning-based test checks synthetic crash data against real pre-crash records
Practical validation of synthetic pre-crash scenarios
-
Hybrid diffusion-flow planner outperforms SOTA on robot tasks
HDFlow: Hierarchical Diffusion-Flow Planning for Long-horizon Tasks
-
Hybrid diffusion-flow planner masters long robot tasks
HDFlow: Hierarchical Diffusion-Flow Planning for Long-horizon Tasks
-
Adaptive homotopy cuts rendezvous miss by 100x with modest extra fuel
Tightly-Coupled Estimation and Guidance for Robust Low-Thrust Rendezvous via Adaptive Homotopy
-
Proxy-residual method lifts closed-loop driving performance
CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policies
-
Unified wrenches fuse position
Autonomous Laparoscope Control through Unified Mechanics-Based Representation of Multimodal Intraoperative Information
-
Flow matching turns normal traffic into safety-critical AV tests
Conditional Flow-VAE for Safety-Critical Traffic Scenario Generation
-
Natural language safety rules become logic specs for robots
From Language to Logic: A Theoretical Architecture for VLM-Grounded Safe Navigation
-
Adaptive thresholds raise F1 scores in multi-task AV tasks
Beyond Fixed Thresholds and Domain-Specific Benchmarks for Explainable Multi-Task Classification in Autonomous Vehicles
-
Framework proposed for open ergonomic modeling in CAD
OPENJ: A Conceptual Framework for Open-Source Digital Human Modeling and Ergonomic Assessment in a CAD Environment
-
Gaussian UOT reduces exactly to SDP optimization
Globally Solving Unbalanced Optimal Transport and Density Control for Gaussian Distributions
-
Pub/sub smart pointer limits reference updates to 0-1 per subscriber
ipc_shared_ptr: A Publish/Subscribe-Aware Smart Pointer for Cross-Process Object Lifetime Management
-
Autoregressive model beats solvers on multi-agent task assignment
ARMATA: Auto-Regressive Multi-Agent Task Assignment
-
Contracts enforce safety by invariance and liveness by refinement in layered control
Safety by Invariance, Liveness through Refinement: Heterogeneous Contract Framework for Co-Design of Layered Control
-
Adaptive per-joint squashing enforces exact actuator constraints in RL
Constraint-Enhanced Reinforcement Learning Based on Dynamic Decoupled Spherical Radial Squashing
-
Hyperdimensional codes set robotic scanner parameters from instructions
Task-Aware Scanning Parameter Configuration for Robotic Inspection Using Vision Language Embeddings and Hyperdimensional Computing
-
Foundation model agents show emergent collaboration in games
Evaluating Generative Models as Interactive Emergent Representations of Human-Like Collaborative Behavior
-
LLM agents show unprompted collaboration in color-matching tasks
Evaluating Generative Models as Interactive Emergent Representations of Human-Like Collaborative Behavior
-
Quadruped matches human teleop on pick-and-place with onboard vision only
SigLoMa: Learning Open-World Quadrupedal Loco-Manipulation from Ego-Centric Vision
-
RL cuts warehouse makespan by 7.5% and order times by 15%
SOAR: Real-Time Joint Optimization of Order Allocation and Robot Scheduling in Robotic Mobile Fulfillment Systems
-
Reward alignment improves robot video world models
RoboAlign-R1: Distilled Multimodal Reward Alignment for Robot Video World Models
-
Guardrails fix LLM failures on drone swarms
Say the Mission, Execute the Swarm: Agent-Enhanced LLM Reasoning in the Web-of-Drones
-
Learning SLAM beats classics in UAV degraded vision tests
Robust Visual SLAM for UAV Navigation in GPS-Denied and Degraded Environments: A Multi-Paradigm Evaluation and Deployment Study
-
Voxel fusion of dense and instance layers scales open-vocabulary mapping
FUS3DMaps: Scalable and Accurate Open-Vocabulary Semantic Mapping by 3D Fusion of Voxel- and Instance-Level Layers
-
Cable constraint model improves quadrotor payload control
Sensorless State Estimation and Control for Agile Cable-Suspended Payload Transport by Quadrotors
-
Hybrid controller merges planning and control for timed robot tasks
Feasibility-aware Hybrid Control for Motion Planning under Signal Temporal Logics
-
Robotic ontology embeds into SMGI for admissible service updates
From Ontology Conformance to Admissible Reconfiguration: A RoSO/SMGI Adequacy Argument for Robotic Service Governance
-
Three components cut robot timing jitter by 84% while allowing safe customization
Jiao: Bridging Isolation and Customization in Mixed Criticality Robotics
-
Disentangled latents turn human videos into robot demos
Bridging the Embodiment Gap: Disentangled Cross-Embodiment Video Editing
-
VR devices collect human data to train humanoid whole-body skills
BifrostUMI: Bridging Robot-Free Demonstrations and Humanoid Whole-Body Manipulation
-
Decoupled RL-QP control enables zero-shot steerable grasping
Learning Reactive Dexterous Grasping via Hierarchical Task-Space RL Planning and Joint-Space QP Control
-
IMU plus triggered satellite matches cut KITTI error 5.9 times
TACO: Trajectory Aligning Cross-view Optimisation
-
Bayesian torque planning cuts robot jump height error 82 percent
Height Control and Optimal Torque Planning for Jumping With Wheeled-Bipedal Robots
-
Risk-aware DR reshapes contact basins for sampling control
On Surprising Effects of Risk-Aware Domain Randomization for Contact-Rich Sampling-based Predictive Control
-
Adjoint gradients enable memory-efficient control of multi-stable objects
Neural Control: Adjoint Learning Through Equilibrium Constraints
-
RLDX-1 humanoid policy reaches 86.8% success on complex tasks
RLDX-1 Technical Report
-
RLDX-1 reaches 86.8% success on complex humanoid tasks
RLDX-1 Technical Report
-
Continuous neural ODEs cut vehicle path errors 69% in MPPI control
Robust Path Tracking for Vehicles via Continuous-Time Residual Learning: An ICODE-MPPI Approach
-
7B models on Raspberry Pi balance speed and teaching for social robots
Benchmarking Local Language Models for Social Robots using Edge Devices
-
Reconstruction error guides stitched diffusion plans past mode averaging
Refining Compositional Diffusion for Reliable Long-Horizon Planning