Attention on Attention: Architectures for Visual Question Answering (VQA)

Jasdeep Singh, Vincent Ying, Alex Nutkiewicz
3/21/2018
cs.CLcs.AIcs.CV

Official code: SinghJasdeep/Attention-on-Attention-for-VQA (131 stars, MIT, last commit 2018-06-18).

Abstract

Visual Question Answering (VQA) is an increasingly popular topic in deep learning research, requiring coordination of natural language processing and computer vision modules into a single architecture. We build upon the model which placed first in the VQA Challenge by developing thirteen new attention mechanisms and introducing a simplified classifier. We performed 300 GPU hours of extensive hyperparameter and architecture searches and were able to achieve an evaluation score of 64.78%, outperforming the existing state-of-the-art single model's validation score of 63.15%.

View on arXivView PDF

Code Implementations(1)

SinghJasdeep/Attention-on-Attention-for-VQAOfficial100%
MIT

Visual Question Answering Project with state of the art single Model performance.

13118Shell, PythonMar 20, 20188 years agoMIT

Similar Papers

The Quest for Visual Understanding: A Journey Through the Evolution of Visual Question Answering

86% match2025

TensorLayer: A Versatile Library for Efficient Deep Learning Development

86% match2 repos7.4k2017

On the Opportunities of Green Computing: A Survey

84% match2023

Caffe: Convolutional Architecture for Fast Feature Embedding

84% match1 repo34.6k2014

Rethinking the Evaluating Framework for Natural Language Understanding in AI Systems: Language Acquisition as a Core for Future Metrics

84% match2023

VMambaCC: A Visual State Space Model for Crowd Counting

84% match2024

Cite this paper

@article{singh2018attention,
  title  = {Attention on Attention: Architectures for Visual Question Answering (VQA)},
  author = {Jasdeep Singh and Vincent Ying and Alex Nutkiewicz},
  year   = {2018},
  eprint = {1803.07724},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1803.07724}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1803.07724.svg)](https://codeofpaper.com/paper/1803.07724)

Discussion