Pythia v0.1: the Winning Entry to the VQA Challenge 2018

TL;DR

Pythia v0.1, developed by Facebook AI Research, is a modular model that achieved state-of-the-art results in the VQA Challenge 2018 by implementing architectural improvements, data augmentation, and diverse ensembling.

Contribution

The paper introduces a modular re-implementation of the bottom-up top-down model with key enhancements that significantly improve VQA performance.

Findings

Achieved 72.27% accuracy on VQA v2.0 test-std split.

Improved performance from 65.67% to 70.22% with architectural and training modifications.

Ensembling diverse models increased accuracy by 1.31%.

Abstract

This document describes Pythia v0.1, the winning entry from Facebook AI Research (FAIR)'s A-STAR team to the VQA Challenge 2018. Our starting point is a modular re-implementation of the bottom-up top-down (up-down) model. We demonstrate that by making subtle but important changes to the model architecture and the learning rate schedule, fine-tuning image features, and adding data augmentation, we can significantly improve the performance of the up-down model on VQA v2.0 dataset -- from 65.67% to 70.22%. Furthermore, by using a diverse ensemble of models trained with different features and on different datasets, we are able to significantly improve over the 'standard' way of ensembling (i.e. same model with different random seeds) by 1.31%. Overall, we achieve 72.27% on the test-std split of the VQA v2.0 dataset. Our code in its entirety (training, evaluation, data-augmentation, ensembling) and pre-trained models are publicly available at: https://github.com/facebookresearch/pythia