Challenge: Contrastively trained Vision-Language Models exhibit shallow language understanding, manifesting bag-of-words behaviour.
Approach: They propose a vision-free, single-encoder retrieval pipeline to replace traditional text-to-image retrieval paradigm with structured image descriptions.
Outcome: The proposed approach reduces the modality gap and improves compositionality and performance on short and long caption queries.

Similar Papers

SERVAL: Surprisingly Effective Zero-Shot Visual Document Retrieval Powered by Large Vision and Language Models (2025.emnlp-main)

Copied to clipboard

Challenge: Visual Document Retrieval (VDR) relies on text-to-image retrieval using specialized bi-encoders . et al., 2022, 2024, 2021, 2023, 2026, 2030, 2040, 2050, 2060) document retrieval bridges human or artificial agents to the most relevant information, authors say .
Approach: They propose a zero-shot visual document retrieval method that uses bi-encoders . they propose 63.4% nDCG@5 for visual document capture and a reusable semantic proxy .
Outcome: The proposed method surpasses the strongest specialised multi-vector visual document encoder on the ViDoRe-v2 benchmark and scales similarly on MIRACL-VISION with broader multilingual coverage.
Text encoders bottleneck compositionality in contrastive vision-language models (2023.emnlp-main)

Copied to clipboard

Challenge: Existing multimodal models are often unable to reason about simple spatial relations or attribute attachments.
Approach: They first curate CompPrompts, a set of increasingly compositional image captions that VL models should be able to capture . then train text-only recovery probes that aim to reconstruct captions from single-vector text representations produced by several VL model.
Outcome: The proposed model can reconstruct captions from single-vector text representations produced by several models on a broader range of scenes compared to previous models.
Exploring How Generative MLLMs Perceive More Than CLIP with the Same Vision Encoder (2025.acl-long)

Copied to clipboard

Challenge: Recent studies show that CLIP models struggle with visual reasoning tasks . despite the success of Contrastive Language-Image Pretraining, there are still limitations .
Approach: They propose to use a visual encoder to train CLIP-like models for fine-grained visual reasoning tasks.
Outcome: The proposed models outperform CLIP-like encoders in visual reasoning tasks . the study highlights the importance of VLM architectural choices .
Scene-Text Aware Image and Text Retrieval with Dual-Encoder (2022.acl-srw)

Copied to clipboard

Challenge: Existing studies on image and text retrieval using a dual-encoder model have not shown their effectiveness for fast inferences.
Approach: They propose a dual-encoder model that connects vision and language in the same semantic space and integrates scene-text and visual information into a model.
Outcome: The proposed model can interpret scene-text and surrounding visual information better than cross-encoder models.
VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval (2024.acl-long)

Copied to clipboard

Challenge: Existing methods for multimodal retrieval are mostly text-oriented, which lack the capability to process visual information.
Approach: They propose a multi-modal multi-text embedding model VISTA which extends a powerful text encoder with the image understanding capability by introducing visual token embedds.
Outcome: The proposed model achieves superior performance across a variety of multi-modal retrieval tasks in zero-shot and supervised settings.
Multimodal Pretraining Unmasked: A Meta-Analysis and a Unified Framework of Vision-and-Language BERTs (2021.tacl-1)

Copied to clipboard

Challenge: Large-scale pretraining and task-specific fine-tuning are now the standard methodology for many tasks in computer vision and natural language processing.
Approach: They propose to combine two types of vision and language BERTs to create a theoretical framework that can be unified under different theoretical frameworks.
Outcome: The proposed models can be classified into single-stream or dual-stream encoders and are unified under a single theoretical framework.
End-to-End Unsupervised Vision-and-Language Pre-training with Referring Expression Matching (2022.emnlp-main)

Copied to clipboard

Challenge: Existing unsupervised vision-and-language pre-training methods take pre-extracted region-based visual features from external object detectors, which limits flexibility and reduces computational efficiency.
Approach: They propose an unsupervised vision-and-language pre-training task that predicts which patches contain an object referred to in natural language from the encoded visual features.
Outcome: The proposed approach outperforms existing methods and obtains state-of-the-art results on four vision-and-language tasks.
MegaPairs: Massive Data Synthesis for Universal Multimodal Retrieval (2025.acl-long)

Copied to clipboard

Challenge: despite the growing demand for multimodal retrieval, there is a lack of training data.
Approach: They propose a data synthesis method that leverages vision language models and open-domain images to generate high-quality data.
Outcome: The proposed method outperforms baseline models on 70 more datasets and can scale up.
On Advances in Text Generation from Images Beyond Captioning: A Case Study in Self-Rationalization (2022.findings-emnlp)

Copied to clipboard

Challenge: Combining visual modality with pretrained language models has been effective for descriptive tasks such as image captioning.
Approach: They ask: do multimodal models combine visual and visual adapted language models? they find that CLIP image representations and scaling of language models do not consistently improve self-rationalization in multimodal tasks.
Outcome: The proposed model types do not consistently improve self-rationalization in multimodal tasks.
Leaner and Faster: Two-Stage Model Compression for Lightweight Text-Image Retrieval (2022.naacl-main)

Copied to clipboard

Challenge: Existing text-image approaches use pre-trained vision-language representations for text retrieval . however, these models pose non-trivial memory requirements and substantial indexing time .
Approach: They propose a framework to compress large pre-trained dual-encoders for lightweight text-image retrieval.
Outcome: The proposed model performs better on Flickr30K and MSCOCO benchmarks than the original full model on mobile devices.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations