English

Grounding Visual Explanations (Extended Abstract)

Computer Vision and Pattern Recognition 2017-11-20 v1

Abstract

Existing models which generate textual explanations enforce task relevance through a discriminative term loss function, but such mechanisms only weakly constrain mentioned object parts to actually be present in the image. In this paper, a new model is proposed for generating explanations by utilizing localized grounding of constituent phrases in generated explanations to ensure image relevance. Specifically, we introduce a phrase-critic model to refine (re-score/re-rank) generated candidate explanations and employ a relative-attribute inspired ranking loss using "flipped" phrases as negative examples for training. At test time, our phrase-critic model takes an image and a candidate explanation as input and outputs a score indicating how well the candidate explanation is grounded in the image.

Keywords

Cite

@article{arxiv.1711.06465,
  title  = {Grounding Visual Explanations (Extended Abstract)},
  author = {Lisa Anne Hendricks and Ronghang Hu and Trevor Darrell and Zeynep Akata},
  journal= {arXiv preprint arXiv:1711.06465},
  year   = {2017}
}

Comments

Presented at NIPS 2017 Symposium on Interpretable Machine Learning

R2 v1 2026-06-22T22:49:09.769Z