Object Search in Partially-Known Environments via LLM-informed Model-based Planning and Prompt Selection

1. 1Estimate statistics about the likelihood of finding the target object using an LLM
2. 2Combine the estimated statistics with travel costs extracted from the environment map
3. 3Instantiate a model using the combined information to inform planning
4. 4Use the LLM-informed model to select optimal prompts for object search

AI engineers and researchers on a team can benefit from this approach to improve object search efficiency, while product managers can apply this technology to develop more effective robotic search systems

💡 Using LLMs to estimate search statistics can improve object search efficiency in partially-known environments

💡 LLM-informed planning for object search in partially-known environments