Research map / Multi-agent systems and games / Human-AI and goal-conditioned
SPEAR: A Simulator for Photorealistic Embodied AI Research
Mike Roberts, Renhan Wang, Rushikesh Zawar, Rachith Dey-Prakash, Quentin Leboutet, Stephan Richter, Matthias Müller, German Ros, Rui Tang, Stefan Leutenegger, Yannick Hold-Geoffroy, Kalyan Sunkavalli and 1 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| enable both asynchronous operations and shared memory (ablation) | SPEAR camera sensor · 1920x1080 render to Python | FPS: 24.7 -> 56.2 | Improved |
| add shared memory (no async) (ablation) | SPEAR camera sensor · 1920x1080 render to Python | FPS: 24.7 -> 31.7 | Improved |
| add asynchronous operations (no shared mem) (ablation) | SPEAR camera sensor · 1920x1080 render to Python | FPS: 24.7 -> 26.8 | Improved |
| increase rendering latency from 0 to 2 frames (ablation) | SPEAR camera sensor · 1920x1080 render to Python | FPS: 56.2 -> 73.4 | Improved |
| use SPEAR instead of UnrealCV+ | UE rendering to Python · 1920x1080 render to Python | FPS: 3.5 -> 56.2 | Improved |
| use SPEAR instead of AirSim | UE rendering to Python · Cross-simulator render, 0 frames latency | FPS: 2.6 -> 32.3 | Improved |
| use SPEAR instead of CARLA at 2 frames latency | UE rendering to Python · Cross-simulator render, 2 frames latency | FPS: 32.7 -> 37.1 | Improved |
| expose UE reflection system to Python | SPEAR · Exposed UE functionality | UE functions exposed: 747 -> 14,485 | Improved |
Closest papers by meaning
- Learning Interactive Real-World Simulators (ICLR 2024 · similarity 0.86)
- LabUtopia: High-Fidelity Simulation and Hierarchical Benchmark for Scientific Embodied Agents (NeurIPS 2025 · similarity 0.85)
- PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AI (CVPR 2024 · similarity 0.85)
- Image2Sim: Scaling Embodied Navigation via Generative Neural Simulator (NeurIPS 2026 · similarity 0.85)
- Synthesizing Photorealistic and Dynamic Urban Environments for Multimodal Robot Navigation and Collaboration (NeurIPS 2025 · similarity 0.85)
- PRISM: Programming Interactive Scenes from Monocular Images for Embodied Simulation (NeurIPS 2026 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
