pith. sign in

arxiv: 1711.06465 · v1 · pith:N46CGTKYnew · submitted 2017-11-17 · 💻 cs.CV

Grounding Visual Explanations (Extended Abstract)

classification 💻 cs.CV
keywords explanationsimagecandidatemodelexplanationgeneratedgroundingloss
0
0 comments X
read the original 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.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.