General Dynamic Goal Recognition
By: Osher Elhadad, Reuth Mirsky
Potential Business Impact:
Helps robots learn what you want them to do.
Understanding an agent's intent through its behavior is essential in human-robot interaction, interactive AI systems, and multi-agent collaborations. This task, known as Goal Recognition (GR), poses significant challenges in dynamic environments where goals are numerous and constantly evolving. Traditional GR methods, designed for a predefined set of goals, often struggle to adapt to these dynamic scenarios. To address this limitation, we introduce the General Dynamic GR problem - a broader definition of GR - aimed at enabling real-time GR systems and fostering further research in this area. Expanding on this foundation, this paper employs a model-free goal-conditioned RL approach to enable fast adaptation for GR across various changing tasks.
Similar Papers
GRAML: Goal Recognition As Metric Learning
Artificial Intelligence
Teaches computers to guess goals from actions.
General and Efficient Visual Goal-Conditioned Reinforcement Learning using Object-Agnostic Masks
CV and Pattern Recognition
Teaches robots to grab any object without knowing its location.
Improving Generalization in Intent Detection: GRPO with Reward-Based Curriculum Sampling
Computation and Language
Teaches computers to understand new requests better.