Unifying Two-Stream Encoders with Transformers for Cross-Modal Retrieval

Abstract

Most existing cross-modal retrieval methods employ two-stream encoders with different architectures for images and texts, \textit{e.g.}, CNN for images and RNN/Transformer for texts. Such discrepancy in architectures may induce different semantic distribution spaces and limit the interactions between images and texts, and further result in inferior alignment between images and texts. To fill this research gap, inspired by recent advances of Transformers in vision tasks, we propose to unify the encoder architectures with Transformers for both modalities. Specifically, we design a cross-modal retrieval framework purely based on two-stream Transformers, dubbed \textbf{Hierarchical Alignment Transformers (HAT)}, which consists of an image Transformer, a text Transformer, and a hierarchical alignment module. With such identical architectures, the encoders could produce representations with more similar characteristics for images and texts, and make the interactions and alignments between them much easier. Besides, to leverage the rich semantics, we devise a hierarchical alignment scheme to explore multi-level correspondences of different layers between images and texts. To evaluate the effectiveness of the proposed HAT, we conduct extensive experiments on two benchmark datasets, MSCOCO and Flickr30K. Experimental results demonstrate that HAT outperforms SOTA baselines by a large margin. Specifically, on two key tasks, \textit{i.e.}, image-to-text and text-to-image retrieval, HAT achieves 7.6\% and 16.7\% relative score improvement of Recall@1 on MSCOCO, and 4.4\% and 11.6\% on Flickr30k respectively. The code is available at \url{https://github.com/LuminosityX/HAT}.

arXiv commentAccepted at ACM Multimedia 2023

References in corpus (15)

• Adam: A Method for Stochastic Optimization
• Efficient Estimation of Word Representations in Vector Space
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Neural Machine Translation by Jointly Learning to Align and Translate
• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• Microsoft COCO Captions: Data Collection and Evaluation Server
• Unifying Visual-Semantic Embeddings with Multimodal Neural Language Models
• VSE++: Improving Visual-Semantic Embeddings with Hard Negatives
• Dual-Path Convolutional Image-Text Embeddings with Instance Loss
• Visualizing Deep Convolutional Neural Networks Using Natural Pre-Images
• How Does BERT Answer Questions? A Layer-Wise Analysis of Transformer Representations
• Your Negative May not Be True Negative: Boosting Image-Text Matching with False Negative Elimination
• BEVT: BERT Pretraining of Video Transformers
• Neural Storyboard Artist: Visualizing Stories with Coherent Image Sequences
• A Multi-View Embedding Space for Modeling Internet Images, Tags, and their Semantics

Cited by in corpus (4)

• Multi-modal Reference Learning for Fine-grained Text-to-Image Retrieval
• MM-Forecast: A Multimodal Approach to Temporal Event Forecasting with Large Language Models
• Leveraging Weak Cross-Modal Guidance for Coherence Modelling via Iterative Learning
• Reversed in Time: A Novel Temporal-Emphasized Benchmark for Cross-Modal Video-Text Retrieval