Grounding Vision and Language to 3D Masks for Long-Horizon Box Rearrangement

1. 1Use visual observations to generate 3D masks for objects
2. 2Ground natural-language goals to 3D masks for planning
3. 3Apply long-horizon planning to achieve multi-step box rearrangement tasks
4. 4Evaluate the approach using metrics such as success rate and efficiency

This research benefits AI engineers and ML researchers working on computer vision and NLP tasks, as it provides a new approach to grounding vision and language in 3D environments

💡 The proposed method enables more effective planning in 3D environments by leveraging visual observations and natural-language goals

💡 Grounding vision and language to 3D masks for long-horizon box rearrangement