arXiv:2603.04836 · v3
Beyond Text: Text-Image Fusion for Scalable Two-Tower Retrieval
摘要
E-commerce customers make purchase decisions by jointly considering product text and visual information, yet most industrial retrieval systems rely almost exclusively on textual signals. Deployed at scale on CPU-based nearest-neighbor infrastructure, these systems leave substantial visual evidence untapped. In this work, we study unified text-image fusion for two-tower retrieval models at a large-scale e-commerce platform, training on 20M query-item pairs. We demonstrate that domain-specific fine-tuning and modality-specific query alignment between user queries and both product text and image modalities are crucial for effective multimodal retrieval. Building on these insights, we propose a mixture-of-modality-experts fusion architecture with a bilinear interaction network to capture cross-modal complementary information. Our model achieves up to 4.86% improvement in nDCG@1 for desirability and 2.36% for relevance over a strong text-only baseline, while preserving the pre-computable item embedding design required for production deployment.
作者备注:Accepted as a regular paper to the Industry & Government Program of the 2026 IEEE International Conference on Big Data (IEEE BigData 2026), Phoenix, AZ, December 14-17, 2026. 9 pages, 6 figures, 5 tables. v2: accepted version (revised title and text). (c) 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses