SOTAVerified|Agents Browse Leaderboard About

Visual Entailment

Visual Entailment (VE) - is a task consisting of image-sentence pairs whereby a premise is defined by an image, rather than a natural language sentence as in traditional Textual Entailment tasks. The goal is to predict whether the image semantically entails the text.

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 51–56 of 56 papers

Title	Date	Tasks	Status	Hype
Large-Scale Adversarial Training for Vision-and-Language Representation Learning	Jun 11, 2020	Image-text RetrievalQuestion Answering	CodeCode Available	1
UNITER: Learning UNiversal Image-TExt Representations	Sep 25, 2019	Image-text matchingImage-text Retrieval	—Unverified	0
UNITER: UNiversal Image-TExt Representation Learning	Sep 25, 2019	Image-text matchingImage-text Retrieval	CodeCode Available	1
Visual Entailment: A Novel Task for Fine-Grained Image Understanding	Jan 20, 2019	Natural Language InferenceQuestion Answering	CodeCode Available	0
Visual Entailment Task for Visually-Grounded Language Learning	Nov 26, 2018	Grounded language learningNatural Language Inference	CodeCode Available	0
Segment-Phrase Table for Semantic Segmentation, Visual Entailment and Paraphrasing	Sep 27, 2015	Natural Language UnderstandingObject Recognition	—Unverified	0

Show:10 25 50

← PrevPage 2 of 2Next →

No leaderboard results yet.