Leveraging Pretrained Unimodal Models for Efficient Image-Text Retrieval
- Existing approaches to image-text retrieval often require large-scale models, extensive data, and substantial computational resources, limiting their accessibility for smaller research groups. We introduce LiteITR, an efficient self-supervised vision-language model that leverages pretrained unimodal encoders with contrastive learning and self-supervised knowledge distillation. While not reaching state-of-the-art performance, our approach demonstrates reasonable performance on retrieval tasks with dramatically reduced resources, requiring only 3M image-text pairs and costing approximately $20 to train. These findings underscore the potential for designing efficient multimodal retrieval systems that are trainable by researchers with limited resources.
Download full text files
Accepted Manuscript. © 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. DOI of the original publication: https://doi.org/10.1109/IDAACS68557.2025.11322222
Additional Services
Statistics
| Author: | Tim Cares, Adrian PigorsGND, Volker AhlersORCiDGND |
|---|---|
| URN: | urn:nbn:de:bsz:960-opus4-38185 |
| DOI: | https://doi.org/10.25968/opus-3818 |
| DOI original: | https://doi.org/10.1109/IDAACS68557.2025.11322222 |
| ISBN: | 979-8-3315-8045-2 |
| ISSN: | 2770-4254 |
| Parent Title (English): | 2025 IEEE 13th International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS) |
| Publisher: | IEEE |
| Document Type: | Conference Proceeding |
| Language: | English |
| Year of Completion: | 2025 |
| Publishing Institution: | Hochschule Hannover |
| Release Date: | 2026/01/16 |
| Tag: | efficiency; image-text retrieval; knowledge distillation; self-supervised learning; vision-language model |
| GND Keyword: | FreitextsucheGND; Information RetrievalGND |
| First Page: | 741 |
| Last Page: | 746 |
| Institutes: | Fakultät IV - Wirtschaft und Informatik |
| Data|H - Institute for Applied Data Science Hannover | |
| DDC classes: | 004 Informatik |
| Licence (German): |  Urheberrechtlich geschützt |
