2019/08/05 · We propose a novel CQA algorithm called parallel recurrent fusion of image and language (PReFIL). PReFIL first learns bimodal embeddings by fusing question and ...
PReFIL first learns bimodal embeddings by fusing question and image features and then intelligently aggregates these learned embeddings to answer the given.
Our results highlight the importance of training on synthetic data, as was done in LLaVA-NeXT, for achieving strong performance.
This repository provides code and pretrained models for the PReFIL algorithm described in the WACV 2020 Paper: Answering Questions about Data Visualizations ...
Answering Questions about Data Visualizations using Efficient Bimodal Fusion. (Supplementary Materials). Kushal Kafle1. Robik Shrestha1. Brian Price2. Scott ...
A newly proposed visual question answering (VQA) task where an algorithm must answer questions about data visualizations, eg bar charts, pie charts, and line ...
2019/08/05 · Here, we propose a novel CQA algorithm called parallel recurrent fusion of image and language (PReFIL). PReFIL first learns bimodal embeddings ...
2019/08/05 · This work proposes a novel CQA algorithm called parallel recurrent fusion of image and language (PReFIL), which first learns bimodal ...
2020/03/01 · In this paper, we investigate whether visual question answering (VQA) systems trained to answer a question about an image, are able to answer ...
Code for the WACV 2020 paper "Answering Questions about Data Visualizations using Efficient Bimodal Fusion"