Unicoder-VL: A Universal Encoder for Vision and Language by Cross-modal Pre-training
This addresses the challenge of integrating vision and language for AI applications, though it is incremental as it builds on existing cross-lingual pre-training ideas.
The authors tackled the problem of learning joint vision-language representations by proposing Unicoder-VL, a cross-modal pre-trained encoder using Transformer and tasks like masked modeling and matching, achieving state-of-the-art or comparable results on image-text retrieval and visual commonsense reasoning.
We propose Unicoder-VL, a universal encoder that aims to learn joint representations of vision and language in a pre-training manner. Borrow ideas from cross-lingual pre-trained models, such as XLM and Unicoder, both visual and linguistic contents are fed into a multi-layer Transformer for the cross-modal pre-training, where three pre-trained tasks are employed, including Masked Language Modeling (MLM), Masked Object Classification (MOC) and Visual-linguistic Matching (VLM). The first two tasks learn context-aware representations for input tokens based on linguistic and visual contents jointly. The last task tries to predict whether an image and a text describe each other. After pretraining on large-scale image-caption pairs, we transfer Unicoder-VL to caption-based image-text retrieval and visual commonsense reasoning, with just one additional output layer. We achieve state-of-the-art or comparable results on both two tasks and show the powerful ability of the cross-modal pre-training.