Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Embodied Amodal Recognition: Learning to Move to Perceive Objects
Jianwei Yang, Zhile Ren, Mingze Xu, Xinlei Chen, David J. Crandall, Devi Parikh, Dhruv Batra
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 |
|---|---|---|---|
| add movement at test time (random path) to passive-trained model | SP/PP · EAR test set (hard) | Clss-Acc: 90.9 -> 91.1 | Improved |
| train on shortest-path frames instead of only the first frame, test passive (ablation) | Temporal Mask R-CNN · EAR test set (all) | Clss-Acc: 92.9 -> 92.8 | No clear effect |
| add embodiment with learned active path plus fine-tuning (AP/AP) versus passive | Embodied Mask R-CNN · EAR test set (hard) | Clss-Acc: 90.9 -> 92.2 | Improved |
| replace shortest-path moves with learned policy (SP/AP vs SP/SP) | Embodied Mask R-CNN · EAR test set (all) | AMask-IoU: 68.6 -> 68.8 | Improved |
| fine-tune amodal recognition on learned paths (AP/AP vs SP/AP) | Embodied Mask R-CNN · EAR test set (all) | Clss-Acc: 93.3 -> 93.7 | Improved |
| replace GRU with max/average pooling for feature aggregation (ablation) | Temporal Mask R-CNN · EAR test set (hard) | all four metrics over steps | Hurt |
| add optical-flow feature warping to GRU aggregation (GRU/OP) (ablation) | Temporal Mask R-CNN · EAR test set (hard) | all four metrics over steps | Improved |
| warp features only (without GRU) (ablation) | Temporal Mask R-CNN · EAR test set (hard) | all four metrics over steps | No clear effect |
Closest papers by meaning
- Selective Visual Representations Improve Convergence and Generalization for Embodied AI (ICLR 2024 · similarity 0.86)
- EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI (CVPR 2024 · similarity 0.86)
- EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents (ICML 2025 · similarity 0.85)
- An Embodied Generalist Agent in 3D World (ICML 2024 · similarity 0.85)
- Active Open-Vocabulary Recognition: Let Intelligent Moving Mitigate CLIP Limitations (CVPR 2024 · similarity 0.85)
- Vision-Language Embodiment for Monocular Depth Estimation (CVPR 2025 · similarity 0.85)
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.
