archive
Every paper Pith has read. Search by title, abstract, or pith.
2900 papers in cs.RO · page 6
-
UAV dataset lifts tracking success by 53 to 69 points
CosFly-Track: A Large-Scale Multi-Modal Dataset for UAV Visual Tracking via Multi-Constraint Trajectory Optimization
-
Optimization turns robot sensor data into physically consistent trajectories
PRIME: Physically-consistent Robotic Inertial and Motion Estimation for Legged and Humanoid Robots
-
Monocular RGB+IMU matches RGB-D accuracy for indoor scene graphs
Mono-Hydra++: Real-Time Monocular Scene Graph Construction with Multi-Task Learning for 3D Indoor Mapping
-
One demo yields general policy for contact-rich robot tasks
From a Single Demonstration to a General Policy for Contact-Rich Manipulation
-
Averaging viewpoints over future object poses improves moving-object reconstruction
Motion-Uncertainty-Aware Next-Best-View Planning for Moving Object Reconstruction
-
Visual features match 3D performance for 100+ step clay sculpting
Visual Sculpting: Visually-Aligned Planning Representations for Long-Horizon Robot Clay Sculpting
-
CBF safety filter maintains leader visibility in UAV formations
Distributed 3D Leader-Follower Formation Control with Field-of-View Safety via Control Barrier Functions
-
Unified model predicts 3D flows for real-time robot manipulation
RoboFlow4D: A Lightweight Flow World Model Toward Real-Time Flow-Guided Robotic Manipulation
-
Affordance alignment raises VLA robot success rates
AffordVLA: Injecting Affordance Representations into Vision-Language-Action Models via Implicit Feature Alignment
-
Dynamic grouped residuals boost VLA model performance across tasks
DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization
-
Boundary controller damps vibrations in flexible robot arms
Rapid Vibration Suppression and Trajectory Tracking of a Serial Manipulator with Multi-Flexible Links
-
MUSE improves uncertainty estimates in robot localization
MUSE: Multimodal Uncertainty Quantification of State Estimation
-
Survey organizes tactile fusion into datasets and three method pillars
Tactile-based Multimodal Fusion in Embodied Intelligence: A Survey of Vision, Language, and Contact-Driven Paradigms
-
Collision knowledge picks primary adversary for AV crash tests
KG-ASG: Collision-Knowledge-Guided Closed-Loop Adversarial Scenario Generation With Primary-Support Attribution
-
Feed-forward 3D model cuts monocular VINS init to under 1.2 s
Efficient Feature-Free Initialization for Monocular Visual-Inertial Systems Using a Feed-Forward 3D Model
-
Surface extraction reduces 3D robot navigation space by over 80%
Beyond Geometry: Efficient Topologically-Grounded Navigation in Complex 3D Environments
-
Quadruped pairs cooperate on manipulation by decoupling control layers
HCLM: A Hierarchical Framework for Cooperative Loco-Manipulation with Dual Quadrupeds
-
Off-center grasps raise collaborative robot capability by 5.86%
Task Capability Improvement Algorithm for Collaborative Manipulators
-
CLAP cuts rare driving errors by 24% via latent prompts
CLAP: Contrastive Latent-space Prompt Optimization for End-to-end Autonomous Driving
-
VLA driving explanations match scenes less than half the time
Is VLA Reasoning Faithful? Probing Safety of Chain-of-Causation in Autonomous Driving Models
-
Stretch-ICP cuts velocity errors 95% in aggressive lidar scans
Stretch-ICP: A Continuous-Trajectory Registration and Deskewing Algorithm in Scenarios of Aggressive Motions
-
Dual spatial system tops VLN benchmarks
SEDualVLN: A Spatially-Enhanced Dual-System for Vision-Language Navigation
-
Three-stage model generates realistic vehicle-pedestrian crisis data
Generating Realistic Safety-Critical Scenarios for Vehicle-Pedestrian Interactions
-
Behavior events ground autoencoder features in robot policies
Event-Grounded Sparse Autoencoders for Vision-Language-Action Policies
-
Steering VLA hidden states lifts robot success over 40 percent
Contrastive Conceptor Activation Steering (COAST): Unlocking Vision-Language-Action Models through Hidden States
-
Drone-sensor fusion hits 85.7% accuracy on casualty action classification
ATRACT: A Trustworthy Robotic Autonomous system to support Casualty Triage
-
Dense labels from existing demos boost robot success
How to Instruct Your Robot: Dense Language Annotations Power Robot Policy Learning
-
Self-supervised learning estimates part poses without symmetry labels
Generalizable and Actionable Parts Pose Estimation with Symmetry Annotation-Free Learning Strategy
-
NORM-Nav turns spoken rules into robot paths
NORM-Nav: Zero-Shot Mobile Robot Navigation with Natural Language Behavioral Constraints
-
Metacognitive controller raises robot goal success from 23 to 30 percent
MORN: Metacognitive Object-Goal Regulation for Resource-Rational Long-Horizon Navigation
-
CBF rewards raise RL performance and stability in vehicle intersections
Beyond Safety Filtering: Control Barrier Function-Informed Reinforcement Learning for Connected and Automated Vehicles
-
Subgoal conditioning boosts robot diffusion policy performance
SADP: Subgoal-Aware Diffusion Policy for Explainable Robots Learned from Foundation Model Generated Demonstrations
-
Proximal tendon loop cuts hysteresis error by 88 percent
SSTL: Self-Sensing Tendon Loop for Hysteresis Modeling and Compensation in Tendon-Sheath Mechanisms
-
Graph plan computed first guides diffusion for long tasks
Plan First, Diffuse Later: Extrinsic Graph Guidance for Long-Horizon Diffusion Planning
-
LLM agent reaches 96% success avoiding pedestrians in driving sim
Pedestrian-Aware LLM-Driven Behavioral Planning for Autonomous Vehicles
-
VLM emotion reading aligns with humans and wins user preference
"I'm Not Mad, Just Focused'': Understanding Human Emotions in Human-Robot Collaboration
-
Toolkit unifies egocentric recording across six device types
EgoKit: Towards Unified Low-Cost Egocentric Data Collection with Heterogeneous Devices
-
LACE boosts zero-shot robot transfer by 65% via human feature alignment
LACE: Latent Visual Representation for Cross-Embodiment Learning
-
Safety framework halves catastrophic failures in driving AI
DriveSafer: End-to-End Autonomous Driving with Safety Guidance
-
Ensemble averaging and uncertainty penalties lift MPC sample efficiency
EfficientTDMPC: Improved MPC Objectives for Sample-Efficient Continuous Control
-
The paper presents a Bayesian Network method to update spatial belief maps from…
Bayesian Networks for Path-Based Sensors: Gathering Information and Path Planning in Communication Denied Environments
-
Parallel rollouts from policy library certify finite-horizon safety
Policy Library CBF: Finite-Horizon Safety at Runtime via Parallel Rollouts
-
AI agent generates lab protocols from prompts at 97% first-try success
From Prompts to Protocols: An AI Agent for Laboratory Automation
-
Sub-$1000 robot reaches floor to counter with built-in safety
Nori Bot: A Sub-$1,000 Floor-to-Counter Mobile Manipulator
-
Hybrid planner adapts robots to new assembly shapes with few examples
EUPHORIA: Efficient Universal Planning via Hybrid Optimization for Robust Industrial Robotic Assembly
-
Sensorimotor rules alone produce collective animal motion
A Mechanistic Model for Collective Motion from Sensorimotor Regularities
-
Reactive model explains human planning failures when capacity drops
No Plan, Yet Human: A Reactive Robotics Model Predicts Human Planning Failures on a Clinical Task
-
Transformer extracts continuous 3D geometry from unposed photos
IVGT: Implicit Visual Geometry Transformer for Neural Scene Representation
-
Implicit transformer builds continuous 3D geometry from unposed views
IVGT: Implicit Visual Geometry Transformer for Neural Scene Representation
-
DexJoCo benchmark reveals dexterous policy shortfalls on tool-use tasks
DexJoCo: A Benchmark and Toolkit for Task-Oriented Dexterous Manipulation on MuJoCo