SOTAVerified

Visual Commonsense Tests

Predict 5 property types (color, shape, material, size, and visual co-occurrence) for over 5000 subjects.

Papers

Showing 12 of 2 papers

TitleStatusHype
Visual Commonsense in Pretrained Unimodal and Multimodal ModelsCode1
Z-LaVI: Zero-Shot Language Solver Fueled by Visual ImaginationCode0
Show:102550

No leaderboard results yet.