![[2602.11575] ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.11575] ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles
Summary
The paper presents ReaDy-Go, a novel simulation pipeline that enhances visual navigation in dynamic environments by integrating 3D Gaussian Splatting with animatable human obstacles, improving navigation performance in real-world applications.
Why It Matters
As robotics and AI applications increasingly operate in dynamic environments, effective navigation strategies are crucial. ReaDy-Go addresses the challenges of sim-to-real transfer and obstacle navigation, paving the way for more robust robotic systems in everyday settings.
Key Takeaways
- ReaDy-Go synthesizes photorealistic dynamic scenarios for navigation training.
- The pipeline integrates static and dynamic elements to enhance simulation realism.
- It demonstrates improved performance in both simulated and real-world environments.
- Zero-shot deployment showcases its generalization capabilities in unseen settings.
- The approach addresses the sim-to-real gap effectively, crucial for practical applications.
Computer Science > Robotics arXiv:2602.11575 (cs) [Submitted on 12 Feb 2026 (v1), last revised 14 Feb 2026 (this version, v2)] Title:ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles Authors:Seungyeon Yoo, Youngseok Jang, Dabin Kim, Youngsoo Han, Seungwoo Jung, H. Jin Kim View a PDF of the paper titled ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles, by Seungyeon Yoo and 5 other authors View PDF HTML (experimental) Abstract:Visual navigation models often struggle in real-world dynamic environments due to limited robustness to the sim-to-real gap and the difficulty of training policies tailored to target deployment environments (e.g., households, restaurants, and factories). Although real-to-sim navigation simulation using 3D Gaussian Splatting (GS) can mitigate these challenges, prior GS-based works have considered only static scenes or non-photorealistic human obstacles built from simulator assets, despite the importance of safe navigation in dynamic environments. To address these issues, we propose ReaDy-Go, a novel real-to-sim simulation pipeline that synthesizes photorealistic dynamic scenarios in target environments by augmenting a reconstructed static GS scene with dynamic human GS obstacles, and trains navigation policies using the generated datasets. The pipeline provides three key contributions: (1) a dynamic...