TensorFlow implementation of the CNN-LSTM, Relation Network and text-only baselines for the paper "FigureQA: An Annotated Figure Dataset for Visual Reasoning"
microsoft deep-learning tensorflow python3 neural-networks vqa relation-network relational-reasoning visual-question-answering figure-analysis
-
Updated
Feb 22, 2018 - Python