sim-to-real

The paper presents NSF-HRPT, a framework that combines a Neural Semantic Field with a Hierarchical Risk Perception Tree for quantitative risk assessment in safety-critical autonomous driving scenarios, achieving state-of-the-art performance on synthetic benchmarks and near-state-of-the-art results on real-world datasets.

ReferTrack introduces a referring-then-tracking paradigm for embodied visual tracking, achieving state-of-the-art performance on EVT-Bench with single-view success rates up to 89.4%, and demonstrating robust sim-to-real transfer on legged and humanoid robots.

This paper introduces an affordable real-world benchmark platform for reinforcement learning in AIoT systems, using video games to measure the Sim-to-Real gap and demonstrating significant performance degradation when transferring simulation-trained agents to the real world.

This paper introduces a predictive formulation for deep reinforcement learning that augments the state space with future reference horizons to enable anticipatory control for trajectory tracking. Simulation results show significant error reduction, though zero-shot transfer to physical hardware reveals a sim-to-real gap.

Image2Sim is a neural simulation framework that creates high-fidelity interactive environments from RGB-D images, enabling scalable training for embodied navigation agents. It generates nearly 20K scenes and over 10 million training samples, showing strong benchmark improvements and effective real-world zero-shot transfer.

Warp RL replaces additive residual corrections in reinforcement learning with an invertible, state-conditioned transformation of the base policy's action distribution using monotonic rational-quadratic spline flows, enabling adaptation of distribution shape, scale, and geometry under dynamics shifts. It matches or outperforms residual correction in ManiSkill3 manipulation tasks and achieves 30% faster task completion in a real robot peg-insertion task.

ASPIRE is a continual learning system that autonomously develops and refines robot control programs through iterative exploration, achieving significant improvements in manipulation and household tasks while enabling sim-to-real transfer.

Building a fault-tolerant RL octocopter from scratch, using MuJoCo simulation and PPO training with domain randomization to handle motor failures. The project aims to directly command motors via RL policy without PID loops, focusing on six failure classes.

General Intuition raised $320M at a $2.3B valuation to develop AI agents trained on video game action labels, demonstrating a single model that can play games and control real-world robots with minimal fine-tuning.

This paper proposes a reinforcement learning-driven adaptive sim-to-real alignment method for vibration-based bearing health monitoring, addressing data scarcity and heterogeneous fault-type gaps via proximal policy optimization.

This paper describes the prizewinning solution for the LeHome Challenge at ICRA 2026, where a two-armed robot learns to fold various garments using a novel RL approach with a self-contained value function, asynchronous training, and heavy sim-to-real augmentation.

Play2Perfect is a reinforcement learning framework that uses playful interaction with diverse objects to learn general manipulation skills, then fine-tunes for precise assembly tasks, achieving 33x sample efficiency and zero-shot sim-to-real transfer on tight insertions.

This blog post walks through using the Strands Robots SDK to integrate Hugging Face Hub datasets with physical robot hardware via LeRobot, enabling a single agent loop from data recording to deployment on real robots.

An object-centric residual reinforcement learning framework enhances zero-shot sim-to-real transfer for vision-language-action models, improving success rates from 42% to 76% on manipulation tasks without real-world training.

This paper formalizes the sim-to-real gap for foundation model agents as a Markov Decision Process problem, proposing a unified research agenda to adapt classical solutions like domain randomization for improving agent robustness and reliability in real-world deployment.

NVIDIA released the Anchor Lab dataset on Hugging Face, containing real-world robotics measurements for calibrating simulation to enable zero-shot sim-to-real deployment.

Dream.exe proposes an evaluation framework that uses robotic manipulation tasks to assess video generation models' understanding of physical reality, finding that visual quality does not predict executable motion accuracy.

GRAIL generates diverse humanoid manipulation and locomotion data using 3D assets and video foundation models, enabling effective sim-to-real transfer for humanoid robot control with high real-world success rates.

This paper proposes a policy-neutral execution and measurement layer to bridge the sim-to-real gap in reinforcement learning-based industrial dispatching, enabling structured attribution of execution errors and improving reliability and interpretability.

This review reframes industrial visual sim-to-real as a domain-gap problem organized by prior availability, distinguishing CAD-guided, CAD-unavailable, and boundary-prior settings to connect CAD-based detection and 6D pose-estimation literature with industrial anomaly and surface-inspection literature.