Cross-modal image-text retrieval is challenging because of the diverse possible associations between content from different modalities. Traditional methods learn a single-vector embedding to represent semantics of each sample, but struggle to capture nuanced and diverse relationships that can exist across modalities. Set-based approaches, which represent each sample with multiple embeddings, offer a promising alternative, as they can capture richer and more diverse relationships. In this paper, we show that, despite their promise, these set-based representations continue to face issues including sparse supervision and set collapse, which limits their effectiveness. To address these challenges, we propose Maximal Pair Assignment Similarity to optimize one-to-one matching between embedding sets which preserve semantic diversity within the set. We also introduce two loss functions to further enhance the representations: Global Discriminative Loss to enhance distinction among embeddings, and Intra-Set Divergence Loss to prevent collapse within each set. Our method achieves state-of-the-art performance on MS-COCO and Flickr30k without relying on external data.
Traditional Retrieval-Augmented Generation (RAG) systems employ brute-force inner product search to retrieve the top-k most similar documents, then combined with the user query and passed to a language model. This allows the model to access external knowledge and reduce hallucinations. However, selecting an appropriate k value remains a significant challenge in practical applications: a small k may fail to retrieve sufficient information, while a large k can introduce excessive and irrelevant content. To address this, we propose a hierarchical clustering-based retrieval method that eliminates the need to predefine k. Our approach maintains the accuracy and relevance of system responses while adaptively selecting semantically relevant content. In the experiment stage, we applied our method to a Taiwanese legal dataset with expert-graded queries. The results show that our approach achieves superior performance in expert evaluations and maintains high precision while eliminating the need to predefine k, demonstrating improved accuracy and interpretability in legal text retrieval tasks. Our framework is simple to implement and easily integrates with existing RAG pipelines, making it a practical solution for real-world applications under limited resources.
We present FlagEvalMM, an open-source evaluation framework designed to comprehensively assess multimodal models across a diverse range of vision-language understanding and generation tasks, such as visual question answering, text-to-image/video generation, and image-text retrieval. We decouple model inference from evaluation through an independent evaluation service, thus enabling flexible resource allocation and seamless integration of new tasks and models. Moreover, FlagEvalMM utilizes advanced inference acceleration tools (e.g., vLLM, SGLang) and asynchronous data loading to significantly enhance evaluation efficiency. Extensive experiments show that FlagEvalMM offers accurate and efficient insights into model strengths and limitations, making it a valuable tool for advancing multimodal research. The framework is publicly accessible athttps://github.com/flageval-baai/FlagEvalMM.
Contrastive Language Audio Pretraining (CLAP) is a widely-used method to bridge the gap between audio and text domains. Current CLAP methods enable sound and music retrieval in English, ignoring multilingual spoken content. To address this, we introduce general language audio pretraining (GLAP), which expands CLAP with multilingual and multi-domain abilities. GLAP demonstrates its versatility by achieving competitive performance on standard audio-text retrieval benchmarks like Clotho and AudioCaps, while significantly surpassing existing methods in speech retrieval and classification tasks. Additionally, GLAP achieves strong results on widely used sound-event zero-shot benchmarks, while simultaneously outperforming previous methods on speech content benchmarks. Further keyword spotting evaluations across 50 languages emphasize GLAP's advanced multilingual capabilities. Finally, multilingual sound and music understanding is evaluated across four languages. Checkpoints and Source: https://github.com/xiaomi-research/dasheng-glap.
Dual encoder Vision-Language Models (VLM) such as CLIP are widely used for image-text retrieval tasks. However, those models struggle with compositionality, showing a bag-of-words-like behavior that limits their retrieval performance. Many different training approaches have been proposed to improve the vision-language compositionality capabilities of those models. In comparison, inference-time techniques have received little attention. In this paper, we propose to add simple structure at inference, where, given an image and a caption: i) we divide the image into different smaller crops, ii) we extract text segments, capturing objects, attributes and relations, iii) using a VLM, we find the image crops that better align with text segments obtaining matches, and iv) we compute the final image-text similarity aggregating the individual similarities of the matches. Based on various popular dual encoder VLMs, we evaluate our approach in controlled and natural datasets for VL compositionality. We find that our approach consistently improves the performance of evaluated VLMs without any training, which shows the potential of inference-time techniques. The results are especially good for attribute-object binding as shown in the controlled dataset. As a result of an extensive analysis: i) we show that processing image crops is actually essential for the observed gains in performance, and ii) we identify specific areas to further improve inference-time approaches.
Scene text retrieval has made significant progress with the assistance of accurate text localization. However, existing approaches typically require costly bounding box annotations for training. Besides, they mostly adopt a customized retrieval strategy but struggle to unify various types of queries to meet diverse retrieval needs. To address these issues, we introduce Muti-query Scene Text retrieval with Attention Recycling (MSTAR), a box-free approach for scene text retrieval. It incorporates progressive vision embedding to dynamically capture the multi-grained representation of texts and harmonizes free-style text queries with style-aware instructions. Additionally, a multi-instance matching module is integrated to enhance vision-language alignment. Furthermore, we build the Multi-Query Text Retrieval (MQTR) dataset, the first benchmark designed to evaluate the multi-query scene text retrieval capability of models, comprising four query types and 16k images. Extensive experiments demonstrate the superiority of our method across seven public datasets and the MQTR dataset. Notably, MSTAR marginally surpasses the previous state-of-the-art model by 6.4% in MAP on Total-Text while eliminating box annotation costs. Moreover, on the MQTR benchmark, MSTAR significantly outperforms the previous models by an average of 8.5%. The code and datasets are available at https://github.com/yingift/MSTAR.
Learning medical visual representations from image-report pairs through joint learning has garnered increasing research attention due to its potential to alleviate the data scarcity problem in the medical domain. The primary challenges stem from the lengthy reports that feature complex discourse relations and semantic pathologies. Previous works have predominantly focused on instance-wise or token-wise cross-modal alignment, often neglecting the importance of pathological-level consistency. This paper presents a novel framework PLACE that promotes the Pathological-Level Alignment and enriches the fine-grained details via Correlation Exploration without additional human annotations. Specifically, we propose a novel pathological-level cross-modal alignment (PCMA) approach to maximize the consistency of pathology observations from both images and reports. To facilitate this, a Visual Pathology Observation Extractor is introduced to extract visual pathological observation representations from localized tokens. The PCMA module operates independently of any external disease annotations, enhancing the generalizability and robustness of our methods. Furthermore, we design a proxy task that enforces the model to identify correlations among image patches, thereby enriching the fine-grained details crucial for various downstream tasks. Experimental results demonstrate that our proposed framework achieves new state-of-the-art performance on multiple downstream tasks, including classification, image-to-text retrieval, semantic segmentation, object detection and report generation.
Retrieval-Augmented Generation (RAG) has demonstrated considerable effectiveness in open-domain question answering. However, when applied to heterogeneous documents, comprising both textual and tabular components, existing RAG approaches exhibit critical limitations. The prevailing practice of flattening tables and chunking strategies disrupts the intrinsic tabular structure, leads to information loss, and undermines the reasoning capabilities of LLMs in multi-hop, global queries. To address these challenges, we propose TableRAG, an hybrid framework that unifies textual understanding and complex manipulations over tabular data. TableRAG iteratively operates in four steps: context-sensitive query decomposition, text retrieval, SQL programming and execution, and compositional intermediate answer generation. We also develop HeteQA, a novel benchmark designed to evaluate the multi-hop heterogeneous reasoning capabilities. Experimental results demonstrate that TableRAG consistently outperforms existing baselines on both public datasets and our HeteQA, establishing a new state-of-the-art for heterogeneous document question answering. We release TableRAG at https://github.com/yxh-y/TableRAG/tree/main.
Foundation models represent the most prominent and recent paradigm shift in artificial intelligence. Foundation models are large models, trained on broad data that deliver high accuracy in many downstream tasks, often without fine-tuning. For this reason, models such as CLIP , DINO or Vision Transfomers (ViT), are becoming the bedrock of many industrial AI-powered applications. However, the reliance on pre-trained foundation models also introduces significant security concerns, as these models are vulnerable to adversarial attacks. Such attacks involve deliberately crafted inputs designed to deceive AI systems, jeopardizing their reliability. This paper studies the vulnerabilities of vision foundation models, focusing specifically on CLIP and ViTs, and explores the transferability of adversarial attacks to downstream tasks. We introduce a novel attack, targeting the structure of transformer-based architectures in a task-agnostic fashion. We demonstrate the effectiveness of our attack on several downstream tasks: classification, captioning, image/text retrieval, segmentation and depth estimation. Code available at:https://github.com/HondamunigePrasannaSilva/attack-attention
The parameter-efficient adaptation of the image-text pretraining model CLIP for video-text retrieval is a prominent area of research. While CLIP is focused on image-level vision-language matching, video-text retrieval demands comprehensive understanding at the video level. Three key discrepancies emerge in the transfer from image-level to video-level: vision, language, and alignment. However, existing methods mainly focus on vision while neglecting language and alignment. In this paper, we propose Discrepancy Reduction in Vision, Language, and Alignment (DiscoVLA), which simultaneously mitigates all three discrepancies. Specifically, we introduce Image-Video Features Fusion to integrate image-level and video-level features, effectively tackling both vision and language discrepancies. Additionally, we generate pseudo image captions to learn fine-grained image-level alignment. To mitigate alignment discrepancies, we propose Image-to-Video Alignment Distillation, which leverages image-level alignment knowledge to enhance video-level alignment. Extensive experiments demonstrate the superiority of our DiscoVLA. In particular, on MSRVTT with CLIP (ViT-B/16), DiscoVLA outperforms previous methods by 1.5% in R@1, reaching a final score of 50.5% R@1. The code is available at https://github.com/LunarShen/DsicoVLA.
Medical vision-language alignment through cross-modal contrastive learning shows promising performance in image-text matching tasks, such as retrieval and zero-shot classification. However, conventional cross-modal contrastive learning (CLIP-based) methods suffer from suboptimal visual representation capabilities, which also limits their effectiveness in vision-language alignment. In contrast, although the models pretrained via multimodal masked modeling struggle with direct cross-modal matching, they excel in visual representation. To address this contradiction, we propose ALTA (ALign Through Adapting), an efficient medical vision-language alignment method that utilizes only about 8% of the trainable parameters and less than 1/5 of the computational consumption required for masked record modeling. ALTA achieves superior performance in vision-language matching tasks like retrieval and zero-shot classification by adapting the pretrained vision model from masked record modeling. Additionally, we integrate temporal-multiview radiograph inputs to enhance the information consistency between radiographs and their corresponding descriptions in reports, further improving the vision-language alignment. Experimental evaluations show that ALTA outperforms the best-performing counterpart by over 4% absolute points in text-to-image accuracy and approximately 6% absolute points in image-to-text retrieval accuracy. The adaptation of vision-language models during efficient alignment also promotes better vision and language understanding. Code is publicly available at https://github.com/DopamineLcy/ALTA.
Chinese scene text retrieval is a practical task that aims to search for images containing visual instances of a Chinese query text. This task is extremely challenging because Chinese text often features complex and diverse layouts in real-world scenes. Current efforts tend to inherit the solution for English scene text retrieval, failing to achieve satisfactory performance. In this paper, we establish a Diversified Layout benchmark for Chinese Street View Text Retrieval (DL-CSVTR), which is specifically designed to evaluate retrieval performance across various text layouts, including vertical, cross-line, and partial alignments. To address the limitations in existing methods, we propose Chinese Scene Text Retrieval CLIP (CSTR-CLIP), a novel model that integrates global visual information with multi-granularity alignment training. CSTR-CLIP applies a two-stage training process to overcome previous limitations, such as the exclusion of visual features outside the text region and reliance on single-granularity alignment, thereby enabling the model to effectively handle diverse text layouts. Experiments on existing benchmark show that CSTR-CLIP outperforms the previous state-of-the-art model by 18.82% accuracy and also provides faster inference speed. Further analysis on DL-CSVTR confirms the superior performance of CSTR-CLIP in handling various text layouts. The dataset and code will be publicly available to facilitate research in Chinese scene text retrieval.
Unsupervised keyphrase prediction has gained growing interest in recent years. However, existing methods typically rely on heuristically defined importance scores, which may lead to inaccurate informativeness estimation. In addition, they lack consideration for time efficiency. To solve these problems, we propose ERU-KG, an unsupervised keyphrase generation (UKG) model that consists of an informativeness and a phraseness module. The former estimates the relevance of keyphrase candidates, while the latter generate those candidates. The informativeness module innovates by learning to model informativeness through references (e.g., queries, citation contexts, and titles) and at the term-level, thereby 1) capturing how the key concepts of documents are perceived in different contexts and 2) estimating informativeness of phrases more efficiently by aggregating term informativeness, removing the need for explicit modeling of the candidates. ERU-KG demonstrates its effectiveness on keyphrase generation benchmarks by outperforming unsupervised baselines and achieving on average 89\% of the performance of a supervised model for top 10 predictions. Additionally, to highlight its practical utility, we evaluate the model on text retrieval tasks and show that keyphrases generated by ERU-KG are effective when employed as query and document expansions. Furthermore, inference speed tests reveal that ERU-KG is the fastest among baselines of similar model sizes. Finally, our proposed model can switch between keyphrase generation and extraction by adjusting hyperparameters, catering to diverse application requirements.
We present Distill CLIP (DCLIP), a fine-tuned variant of the CLIP model that enhances multimodal image-text retrieval while preserving the original model's strong zero-shot classification capabilities. CLIP models are typically constrained by fixed image resolutions and limited context, which can hinder their effectiveness in retrieval tasks that require fine-grained cross-modal understanding. DCLIP addresses these challenges through a meta teacher-student distillation framework, where a cross-modal transformer teacher is fine-tuned to produce enriched embeddings via bidirectional cross-attention between YOLO-extracted image regions and corresponding textual spans. These semantically and spatially aligned global representations guide the training of a lightweight student model using a hybrid loss that combines contrastive learning and cosine similarity objectives. Despite being trained on only ~67,500 samples curated from MSCOCO, Flickr30k, and Conceptual Captions-just a fraction of CLIP's original dataset-DCLIP significantly improves image-text retrieval metrics (Recall@K, MAP), while retaining approximately 94% of CLIP's zero-shot classification performance. These results demonstrate that DCLIP effectively mitigates the trade-off between task specialization and generalization, offering a resource-efficient, domain-adaptive, and detail-sensitive solution for advanced vision-language tasks. Code available at https://anonymous.4open.science/r/DCLIP-B772/README.md.
Effective video tokenization is critical for scaling transformer models for long videos. Current approaches tokenize videos using space-time patches, leading to excessive tokens and computational inefficiencies. The best token reduction strategies degrade performance and barely reduce the number of tokens when the camera moves. We introduce grounded video tokenization, a paradigm that organizes tokens based on panoptic sub-object trajectories rather than fixed patches. Our method aligns with fundamental perceptual principles, ensuring that tokenization reflects scene complexity rather than video duration. We propose TrajViT, a video encoder that extracts object trajectories and converts them into semantically meaningful tokens, significantly reducing redundancy while maintaining temporal coherence. Trained with contrastive learning, TrajViT significantly outperforms space-time ViT (ViT3D) across multiple video understanding benchmarks, e.g., TrajViT outperforms ViT3D by a large margin of 6% top-5 recall in average at video-text retrieval task with 10x token deduction. We also show TrajViT as a stronger model than ViT3D for being the video encoder for modern VideoLLM, obtaining an average of 5.2% performance improvement across 6 VideoQA benchmarks while having 4x faster training time and 18x less inference FLOPs. TrajViT is the first efficient encoder to consistently outperform ViT3D across diverse video analysis tasks, making it a robust and scalable solution.
Existing Zero-Shot Composed Image Retrieval (ZS-CIR) methods typically train adapters that convert reference images into pseudo-text tokens, which are concatenated with the modifying text and processed by frozen text encoders in pretrained VLMs or LLMs. While this design leverages the strengths of large pretrained models, it only supervises the adapter to produce encoder-compatible tokens that loosely preserve visual semantics. Crucially, it does not directly optimize the composed query representation to capture the full intent of the composition or to align with the target semantics, thereby limiting retrieval performance, particularly in cases involving fine-grained or complex visual transformations. To address this problem, we propose MLLM-Guided VLM Fine-Tuning with Joint Inference (MVFT-JI), a novel approach that leverages a pretrained multimodal large language model (MLLM) to construct two complementary training tasks using only unlabeled images: target text retrieval taskand text-to-image retrieval task. By jointly optimizing these tasks, our method enables the VLM to inherently acquire robust compositional retrieval capabilities, supported by the provided theoretical justifications and empirical validation. Furthermore, during inference, we further prompt the MLLM to generate target texts from composed queries and compute retrieval scores by integrating similarities between (i) the composed query and candidate images, and (ii) the MLLM-generated target text and candidate images. This strategy effectively combines the VLM's semantic alignment strengths with the MLLM's reasoning capabilities.
Vision-language retrieval (VLR) has attracted significant attention in both academia and industry, which involves using text (or images) as queries to retrieve corresponding images (or text). However, existing methods often neglect the rich visual semantics knowledge of entities, thus leading to incorrect retrieval results. To address this problem, we propose the Entity Visual Description enhanced CLIP (EvdCLIP), designed to leverage the visual knowledge of entities to enrich queries. Specifically, since humans recognize entities through visual cues, we employ a large language model (LLM) to generate Entity Visual Descriptions (EVDs) as alignment cues to complement textual data. These EVDs are then integrated into raw queries to create visually-rich, EVD-enhanced queries. Furthermore, recognizing that EVD-enhanced queries may introduce noise or low-quality expansions, we develop a novel, trainable EVD-aware Rewriter (EaRW) for vision-language retrieval tasks. EaRW utilizes EVD knowledge and the generative capabilities of the language model to effectively rewrite queries. With our specialized training strategy, EaRW can generate high-quality and low-noise EVD-enhanced queries. Extensive quantitative and qualitative experiments on image-text retrieval benchmarks validate the superiority of EvdCLIP on vision-language retrieval tasks.
Remote Sensing Image-Text Retrieval (RSITR) plays a critical role in geographic information interpretation, disaster monitoring, and urban planning by establishing semantic associations between image and textual descriptions. Existing Parameter-Efficient Fine-Tuning (PEFT) methods for Vision-and-Language Pre-training (VLP) models typically adopt symmetric adapter structures for exploring cross-modal correlations. However, the strong discriminative nature of text modality may dominate the optimization process and inhibits image representation learning. The nonnegligible imbalanced cross-modal optimization remains a bottleneck to enhancing the model performance. To address this issue, this study proposes a Representation Discrepancy Bridging (RDB) method for the RSITR task. On the one hand, a Cross-Modal Asymmetric Adapter (CMAA) is designed to enable modality-specific optimization and improve feature alignment. The CMAA comprises a Visual Enhancement Adapter (VEA) and a Text Semantic Adapter (TSA). VEA mines fine-grained image features by Differential Attention (DA) mechanism, while TSA identifies key textual semantics through Hierarchical Attention (HA) mechanism. On the other hand, this study extends the traditional single-task retrieval framework to a dual-task optimization framework and develops a Dual-Task Consistency Loss (DTCL). The DTCL improves cross-modal alignment robustness through an adaptive weighted combination of cross-modal, classification, and exponential moving average consistency constraints. Experiments on RSICD and RSITMD datasets show that the proposed RDB method achieves a 6%-11% improvement in mR metrics compared to state-of-the-art PEFT methods and a 1.15%-2% improvement over the full fine-tuned GeoRSCLIP model.
Multilingual vision-language models promise universal image-text retrieval, yet their social biases remain under-explored. We present the first systematic audit of three public multilingual CLIP checkpoints -- M-CLIP, NLLB-CLIP, and CAPIVARA-CLIP -- across ten languages that vary in resource availability and grammatical gender. Using balanced subsets of \textsc{FairFace} and the \textsc{PATA} stereotype suite in a zero-shot setting, we quantify race and gender bias and measure stereotype amplification. Contrary to the assumption that multilinguality mitigates bias, every model exhibits stronger gender bias than its English-only baseline. CAPIVARA-CLIP shows its largest biases precisely in the low-resource languages it targets, while the shared cross-lingual encoder of NLLB-CLIP transports English gender stereotypes into gender-neutral languages; loosely coupled encoders largely avoid this transfer. Highly gendered languages consistently magnify all measured bias types, but even gender-neutral languages remain vulnerable when cross-lingual weight sharing imports foreign stereotypes. Aggregated metrics conceal language-specific ``hot spots,'' underscoring the need for fine-grained, language-aware bias evaluation in future multilingual vision-language research.
Long videos contain a vast amount of information, making video-text retrieval an essential and challenging task in multimodal learning. However, existing benchmarks suffer from limited video duration, low-quality captions, and coarse annotation granularity, which hinder the evaluation of advanced video-text retrieval methods. To address these limitations, we introduce LoVR, a benchmark specifically designed for long video-text retrieval. LoVR contains 467 long videos and over 40,804 fine-grained clips with high-quality captions. To overcome the issue of poor machine-generated annotations, we propose an efficient caption generation framework that integrates VLM automatic generation, caption quality scoring, and dynamic refinement. This pipeline improves annotation accuracy while maintaining scalability. Furthermore, we introduce a semantic fusion method to generate coherent full-video captions without losing important contextual information. Our benchmark introduces longer videos, more detailed captions, and a larger-scale dataset, presenting new challenges for video understanding and retrieval. Extensive experiments on various advanced embedding models demonstrate that LoVR is a challenging benchmark, revealing the limitations of current approaches and providing valuable insights for future research. We release the code and dataset link at https://github.com/TechNomad-ds/LoVR-benchmark