Research map / Vision-language and multimodal / Robot action and navigation
RoboTHOR: An Open Simulation-to-Real Embodied AI Platform
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add Faster-RCNN object detections to the image input | Image-A3C · RoboTHOR test-dev, Sim-to-Sim, Easy | Success: 55.30 -> 36.36 | Hurt |
| Add Faster-RCNN object detections to the image input | Image-A3C · RoboTHOR test-dev, Sim-to-Sim, Easy | SPL: 38.12 -> 19.89 | Hurt |
| Train in simulation and evaluate on the physical robot rather than in simulation | Image-A3C · RoboTHOR test-dev, Easy | Success: 55.30 -> 33.33 | Hurt |
| Train in simulation and evaluate on the physical robot rather than in simulation | Image-A3C · RoboTHOR test-dev, Easy | SPL: 38.12 -> 3.53 | Hurt |
| Train with 90° FOV and test with 42.5° FOV (ablation) | Image-A3C · RoboTHOR physical test-dev | Success: 16%; 0% for medium and hard | Hurt |
| Translate real images toward simulation images with CycleGAN (ablation) | Image-A3C · RoboTHOR physical test-dev | Success: 0% | Hurt |
| Train on one test-dev scene to memorize its simulated layout, then transfer to the robot | Image-A3C · RoboTHOR test-dev, Hard | Success: 94.12 -> 50 | Hurt |
Closest papers by meaning
- EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI (CVPR 2024 · similarity 0.88)
- PhysBench: Benchmarking and Enhancing Vision-Language Models for Physical World Understanding (ICLR 2025 · similarity 0.88)
- EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents (ICML 2025 · similarity 0.87)
- Learning 3D Persistent Embodied World Models (NeurIPS 2025 · similarity 0.87)
- Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models (NeurIPS 2026 · similarity 0.87)
- WorldSimBench: Towards Video Generation Models as World Simulators (ICML 2025 · similarity 0.87)
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.