Learning to Defer to a Population: A Meta-Learning Approach

The learning to defer (L2D) framework allows autonomous systems to be safe and robust by allocating difficult decisions to a human expert. All existing work on L2D assumes that each expert is well-identified, and if any expert were to change, the system should be re-trained. In this work, we alleviate this constraint, formulating an L2D system that can cope with never-before-seen experts at test-time. We accomplish this by using a meta-learning architecture for the deferral function: given a small context set to identify the currently available expert, the model can quickly adapt its deferral policy. We also employ an attention mechanism that is able to look for points in the context set that are similar to a given test point, leading to an even more precise assessment of the expert's abilities. In the experiments, we demonstrate the usefulness of this architecture on image recognition, traffic sign detection, and skin lesion diagnosis benchmarks.