What BERT Sees: Cross-Modal Transfer for Visual Question Generation
Thomas Scialom, Patrick Bordes, Paul-Alexis Dray, Jacopo Staiano, Patrick Gallinari
Abstract
Pre-trained language models have recently contributed to significant advances in NLP tasks. Recently, multi-modal versions of BERT have been developed, using heavy pre-training relying on vast corpora of aligned textual and image data, primarily applied to classification tasks such as VQA. In this paper, we are interested in evaluating the visual capabilities of BERT out-of-the-box, by avoiding pre-training made on supplementary data. We choose to study Visual Question Generation, a task of great interest for grounded dialog, that enables to study the impact of each modality (as input can be visual and/or textual). Moreover, the generation aspect of the task requires an adaptation since BERT is primarily designed as an encoder. We introduce BERT-gen, a BERT-based architecture for text generation, able to leverage on either mono- or multi- modal representations. The results reported under different configurations indicate an innate capacity for BERT-gen to adapt to multi-modal data and text generation, even with few data available, avoiding expensive pre-training. The proposed model obtains substantial improvements over the state-of-the-art on two established VQG datasets.- Anthology ID:
- 2020.inlg-1.39
- Original:
- 2020.inlg-1.39v1
- Version 2:
- 2020.inlg-1.39v2
- Volume:
- Proceedings of the 13th International Conference on Natural Language Generation
- Month:
- December
- Year:
- 2020
- Address:
- Dublin, Ireland
- Editors:
- Brian Davis, Yvette Graham, John Kelleher, Yaji Sripada
- Venue:
- INLG
- SIG:
- SIGGEN
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 327–337
- Language:
- URL:
- https://aclanthology.org/2020.inlg-1.39
- DOI:
- 10.18653/v1/2020.inlg-1.39
- Cite (ACL):
- Thomas Scialom, Patrick Bordes, Paul-Alexis Dray, Jacopo Staiano, and Patrick Gallinari. 2020. What BERT Sees: Cross-Modal Transfer for Visual Question Generation. In Proceedings of the 13th International Conference on Natural Language Generation, pages 327–337, Dublin, Ireland. Association for Computational Linguistics.
- Cite (Informal):
- What BERT Sees: Cross-Modal Transfer for Visual Question Generation (Scialom et al., INLG 2020)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-3/2020.inlg-1.39.pdf
- Data
- MS COCO, VQG, Visual Question Answering