An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information
This addresses sampling bias in unsupervised multimodal learning for researchers, but it is incremental as it builds on existing document-level structural approaches.
The paper tackles unsupervised image-sentence matching by proposing a new sampling strategy to reduce bias and a Transformer-based model to capture fine-grained features, showing effectiveness in learning aligned multimodal representations.
In this paper, we focus on the problem of unsupervised image-sentence matching. Existing research explores to utilize document-level structural information to sample positive and negative instances for model training. Although the approach achieves positive results, it introduces a sampling bias and fails to distinguish instances with high semantic similarity. To alleviate the bias, we propose a new sampling strategy to select additional intra-document image-sentence pairs as positive or negative samples. Furthermore, to recognize the complex pattern in intra-document samples, we propose a Transformer based model to capture fine-grained features and implicitly construct a graph for each document, where concepts in a document are introduced to bridge the representation learning of images and sentences in the context of a document. Experimental results show the effectiveness of our approach to alleviate the bias and learn well-aligned multimodal representations.