Welcome to the VQA Challenge 2019!


We are pleased to announce the Visual Question Answering (VQA) Challenge 2019. Given an image and a natural language question about the image, the task is to provide an accurate natural language answer. The challenge will be hosted on EvalAI and links to Submission and Leaderboard pages will be posted soon.

The VQA v2.0 train, validation and test sets, containing more than 250K images and 1.1M questions, are available on the download page. All questions are annotated with 10 concise, open-ended answers each. Annotations on the training and validation sets are publicly available.

VQA Challenge 2019 is the fourth edition of the VQA Challenge. Previous three versions of the VQA Challenge were organized in past three years, and the results were announced at VQA Challenge Workshop in CVPR 2018, CVPR 2017 and CVPR 2016. More details about past challenges can be found here: VQA Challenge 2018, VQA Challenge 2017 and VQA Challenge 2016.

Answers to some common questions about the challenge can be found in the FAQ section.


Jan 28, 2019 VQA Challenge 2019 will be launched!
Jun, 2019 Winners' announcement at the VQA and Dialog Workshop, CVPR 2019.

Tools and Instructions

We provide API support for the VQA annotations and evaluation code. To download the VQA API, please visit our GitHub repository. For an overview of how to use the API, please visit the download page and consult the section entitled VQA API. To obtain API support for COCO images, please visit the COCO download page. To obtain API support for abstract scenes, please visit the GitHub repository.

For challenge related questions, please contact visualqa@gmail.com. In case of technical questions related to EvalAI, please post on EvalAI's mailing list.


Ayush Shrivastava
Georgia Tech

Karan Desai
Georgia Tech

Yash Goyal
Georgia Tech

Aishwarya Agrawal
Georgia Tech

Dhruv Batra
Georgia Tech / Facebook AI Research

Devi Parikh
Georgia Tech / Facebook AI Research