A Transformer-based Cross-modal Fusion Model with Adversarial Training for VQA Challenge 2021
arXiv:2106.13033
Abstract
In this paper, inspired by the successes of visionlanguage pre-trained models and the benefits from training with adversarial attacks, we present a novel transformerbased cross-modal fusion modeling by incorporating the both notions for VQA challenge 2021. Specifically, the proposed model is on top of the architecture of VinVL model [19], and the adversarial training strategy [4] is applied to make the model robust and generalized. Moreover, two implementation tricks are also used in our system to obtain better results. The experiments demonstrate that the novel framework can achieve 76.72% on VQAv2 test-std set.
arXiv commentCVPR 2021 Workshop: Visual Question Answering (VQA) Challenge
References in corpus (4)
- VisualBERT: A Simple and Performant Baseline for Vision and Language
- Large-Scale Adversarial Training for Vision-and-Language Representation Learning
- LXMERT: Learning Cross-Modality Encoder Representations from Transformers
- ERNIE-ViL: Knowledge Enhanced Vision-Language Representations Through Scene Graph