Using nearest neighbors to improve both human and AI accuracy

We just finished a paper named PCNN. The project website is live here.

We present a new class of nearest-neighbor explanations (called PCNN) and show a novel utility of the XAI method: To improve predictions of a frozen, pretrained classifier C . Our method consistently improves fine-grained image classification accuracy on CUB-200, Cars-196, and Dogs-120. Also, a human study finds that showing layusers our PCNN explanations improves their decision accuracy over showing only the top-1 class examples (as in prior work).