This paper introduces the ColorSwap dataset, designed to assess and improve the proficiency of multimodal models in matching objects with their colors. The dataset is comprised of 2,000 unique image-caption pairs, grouped into 1,000 examples. Each example includes a caption-image pair, along with a ``color-swapped'' pair. We follow the Winoground schema: the two captions in an example have the same words, but the color words have been rearranged to modify different objects. The dataset was created through a novel blend of automated caption and image generation with humans in the loop. We evaluate image-text matching (ITM) and visual language models (VLMs) and find that even the latest ones are still not robust at this task. GPT-4V and LLaVA score 72% and 42% on our main VLM metric, although they may improve with more advanced prompting techniques. On the main ITM metric, contrastive models such as CLIP and SigLIP perform close to chance (at 12% and 30%, respectively), although the non-contrastive BLIP ITM model is stronger (87%). We also find that finetuning on fewer than 2,000 examples yields significant performance gains on this out-of-distribution word-order understanding task. The dataset is here: https://github.com/Top34051/colorswap and here: https://huggingface.co/datasets/stanfordnlp/colorswap.
The dominant probing approaches rely on the zero-shot performance of image-text matching tasks to gain a finer-grained understanding of the representations learned by recent multimodal image-language transformer models. The evaluation is carried out on carefully curated datasets focusing on counting, relations, attributes, and others. This work introduces an alternative probing strategy called guided masking. The proposed approach ablates different modalities using masking and assesses the model's ability to predict the masked word with high accuracy. We focus on studying multimodal models that consider regions of interest (ROI) features obtained by object detectors as input tokens. We probe the understanding of verbs using guided masking on ViLBERT, LXMERT, UNITER, and VisualBERT and show that these models can predict the correct verb with high accuracy. This contrasts with previous conclusions drawn from image-text matching probing techniques that frequently fail in situations requiring verb understanding. The code for all experiments will be publicly available https://github.com/ivana-13/guided_masking.
Current large vision-language models (VLMs) often encounter challenges such as insufficient capabilities of a single visual component and excessively long visual tokens. These issues can limit the model's effectiveness in accurately interpreting complex visual information and over-lengthy contextual information. Addressing these challenges is crucial for enhancing the performance and applicability of VLMs. This paper proposes the use of ensemble experts technique to synergizes the capabilities of individual visual encoders, including those skilled in image-text matching, OCR, image segmentation, etc. This technique introduces a fusion network to unify the processing of outputs from different visual experts, while bridging the gap between image encoders and pre-trained LLMs. In addition, we explore different positional encoding schemes to alleviate the waste of positional encoding caused by lengthy image feature sequences, effectively addressing the issue of position overflow and length limitations. For instance, in our implementation, this technique significantly reduces the positional occupancy in models like SAM, from a substantial 4096 to a more efficient and manageable 64 or even down to 1. Experimental results demonstrate that VLMs with multiple experts exhibit consistently superior performance over isolated visual encoders and mark a significant performance boost as more experts are integrated. We have open-sourced the training code used in this report. All of these resources can be found on our project website.
Image-text matching aims to find matched cross-modal pairs accurately. While current methods often rely on projecting cross-modal features into a common embedding space, they frequently suffer from imbalanced feature representations across different modalities, leading to unreliable retrieval results. To address these limitations, we introduce a novel Feature Enhancement Module that adaptively aggregates single-modal features for more balanced and robust image-text retrieval. Additionally, we propose a new loss function that overcomes the shortcomings of original triplet ranking loss, thereby significantly improving retrieval performance. The proposed model has been evaluated on two public datasets and achieves competitive retrieval performance when compared with several state-of-the-art models. Implementation codes can be found here.
Vision-Language Models (VLMs), pre-trained on large-scale datasets, have shown impressive performance in various visual recognition tasks. This advancement paves the way for notable performance in Zero-Shot Egocentric Action Recognition (ZS-EAR). Typically, VLMs handle ZS-EAR as a global video-text matching task, which often leads to suboptimal alignment of vision and linguistic knowledge. We propose a refined approach for ZS-EAR using VLMs, emphasizing fine-grained concept-description alignment that capitalizes on the rich semantic and contextual details in egocentric videos. In this paper, we introduce GPT4Ego, a straightforward yet remarkably potent VLM framework for ZS-EAR, designed to enhance the fine-grained alignment of concept and description between vision and language. Extensive experiments demonstrate GPT4Ego significantly outperforms existing VLMs on three large-scale egocentric video benchmarks, i.e., EPIC-KITCHENS-100 (33.2%, +9.4%), EGTEA (39.6%, +5.5%), and CharadesEgo (31.5%, +2.6%).
Customizable keyword spotting (KWS) in continuous speech has attracted increasing attention due to its real-world application potential. While contrastive learning (CL) has been widely used to extract keyword representations, previous CL approaches all operate on pre-segmented isolated words and employ only audio-text representations matching strategy. However, for KWS in continuous speech, co-articulation and streaming word segmentation can easily yield similar audio patterns for different texts, which may consequently trigger false alarms. To address this issue, we propose a novel CL with Audio Discrimination (CLAD) approach to learning keyword representation with both audio-text matching and audio-audio discrimination ability. Here, an InfoNCE loss considering both audio-audio and audio-text CL data pairs is employed for each sliding window during training. Evaluations on the open-source LibriPhrase dataset show that the use of sliding-window level InfoNCE loss yields comparable performance compared to previous CL approaches. Furthermore, experiments on the continuous speech dataset LibriSpeech demonstrate that, by incorporating audio discrimination, CLAD achieves significant performance gain over CL without audio discrimination. Meanwhile, compared to two-stage KWS approaches, the end-to-end KWS with CLAD achieves not only better performance, but also significant speed-up.
Recently, image-text matching has attracted more and more attention from academia and industry, which is fundamental to understanding the latent correspondence across visual and textual modalities. However, most existing methods implicitly assume the training pairs are well-aligned while ignoring the ubiquitous annotation noise, a.k.a noisy correspondence (NC), thereby inevitably leading to a performance drop. Although some methods attempt to address such noise, they still face two challenging problems: excessive memorizing/overfitting and unreliable correction for NC, especially under high noise. To address the two problems, we propose a generalized Cross-modal Robust Complementary Learning framework (CRCL), which benefits from a novel Active Complementary Loss (ACL) and an efficient Self-refining Correspondence Correction (SCC) to improve the robustness of existing methods. Specifically, ACL exploits active and complementary learning losses to reduce the risk of providing erroneous supervision, leading to theoretically and experimentally demonstrated robustness against NC. SCC utilizes multiple self-refining processes with momentum correction to enlarge the receptive field for correcting correspondences, thereby alleviating error accumulation and achieving accurate and stable corrections. We carry out extensive experiments on three image-text benchmarks, i.e., Flickr30K, MS-COCO, and CC152K, to verify the superior robustness of our CRCL against synthetic and real-world noisy correspondences.
When exploring the development of Artificial General Intelligence (AGI), a critical task for these models involves interpreting and processing information from multiple image inputs. However, Large Multimodal Models (LMMs) encounter two issues in such scenarios: (1) a lack of fine-grained perception, and (2) a tendency to blend information across multiple images. We first extensively investigate the capability of LMMs to perceive fine-grained visual details when dealing with multiple input images. The research focuses on two aspects: first, image-to-image matching (to evaluate whether LMMs can effectively reason and pair relevant images), and second, multi-image-to-text matching (to assess whether LMMs can accurately capture and summarize detailed image information). We conduct evaluations on a range of both open-source and closed-source large models, including GPT-4V, Gemini, OpenFlamingo, and MMICL. To enhance model performance, we further develop a Contrastive Chain-of-Thought (CoCoT) prompting approach based on multi-input multimodal models. This method requires LMMs to compare the similarities and differences among multiple image inputs, and then guide the models to answer detailed questions about multi-image inputs based on the identified similarities and differences. Our experimental results showcase CoCoT's proficiency in enhancing the multi-image comprehension capabilities of large multimodal models.
In recent years, foundation models (FMs) have solidified their role as cornerstone advancements in the deep learning domain. By extracting intricate patterns from vast datasets, these models consistently achieve state-of-the-art results across a spectrum of downstream tasks, all without necessitating extensive computational resources. Notably, MedCLIP, a vision-language contrastive learning-based medical FM, has been designed using unpaired image-text training. While the medical domain has often adopted unpaired training to amplify data, the exploration of potential security concerns linked to this approach hasn't kept pace with its practical usage. Notably, the augmentation capabilities inherent in unpaired training also indicate that minor label discrepancies can result in significant model deviations. In this study, we frame this label discrepancy as a backdoor attack problem. We further analyze its impact on medical FMs throughout the FM supply chain. Our evaluation primarily revolves around MedCLIP, emblematic of medical FM employing the unpaired strategy. We begin with an exploration of vulnerabilities in MedCLIP stemming from unpaired image-text matching, termed BadMatch. BadMatch is achieved using a modest set of wrongly labeled data. Subsequently, we disrupt MedCLIP's contrastive learning through BadDist-assisted BadMatch by introducing a Bad-Distance between the embeddings of clean and poisoned data. Additionally, combined with BadMatch and BadDist, the attacking pipeline consistently fends off backdoor assaults across diverse model designs, datasets, and triggers. Also, our findings reveal that current defense strategies are insufficient in detecting these latent threats in medical FMs' supply chains.
Multimodal Entity Linking (MEL) aims at linking ambiguous mentions with multimodal information to entity in Knowledge Graph (KG) such as Wikipedia, which plays a key role in many applications. However, existing methods suffer from shortcomings, including modality impurity such as noise in raw image and ambiguous textual entity representation, which puts obstacles to MEL. We formulate multimodal entity linking as a neural text matching problem where each multimodal information (text and image) is treated as a query, and the model learns the mapping from each query to the relevant entity from candidate entities. This paper introduces a dual-way enhanced (DWE) framework for MEL: (1) our model refines queries with multimodal data and addresses semantic gaps using cross-modal enhancers between text and image information. Besides, DWE innovatively leverages fine-grained image attributes, including facial characteristic and scene feature, to enhance and refine visual features. (2)By using Wikipedia descriptions, DWE enriches entity semantics and obtains more comprehensive textual representation, which reduces between textual representation and the entities in KG. Extensive experiments on three public benchmarks demonstrate that our method achieves state-of-the-art (SOTA) performance, indicating the superiority of our model. The code is released on https://github.com/season1blue/DWE
Vision-language pre-training (VLP) models demonstrate impressive abilities in processing both images and text. However, they are vulnerable to multi-modal adversarial examples (AEs). Investigating the generation of high-transferability adversarial examples is crucial for uncovering VLP models' vulnerabilities in practical scenarios. Recent works have indicated that leveraging data augmentation and image-text modal interactions can enhance the transferability of adversarial examples for VLP models significantly. However, they do not consider the optimal alignment problem between dataaugmented image-text pairs. This oversight leads to adversarial examples that are overly tailored to the source model, thus limiting improvements in transferability. In our research, we first explore the interplay between image sets produced through data augmentation and their corresponding text sets. We find that augmented image samples can align optimally with certain texts while exhibiting less relevance to others. Motivated by this, we propose an Optimal Transport-based Adversarial Attack, dubbed OT-Attack. The proposed method formulates the features of image and text sets as two distinct distributions and employs optimal transport theory to determine the most efficient mapping between them. This optimal mapping informs our generation of adversarial examples to effectively counteract the overfitting issues. Extensive experiments across various network architectures and datasets in image-text matching tasks reveal that our OT-Attack outperforms existing state-of-the-art methods in terms of adversarial transferability.
The age estimation task aims to predict the age of an individual by analyzing facial features in an image. The development of age estimation can improve the efficiency and accuracy of various applications (e.g., age verification and secure access control, etc.). In recent years, contrastive language-image pre-training (CLIP) has been widely used in various multimodal tasks and has made some progress in the field of age estimation. However, existing CLIP-based age estimation methods require high memory usage (quadratic complexity) when globally modeling images, and lack an error feedback mechanism to prompt the model about the quality of age prediction results. To tackle the above issues, we propose a novel CLIP-driven Image-Language Fusion for Correcting Inverse Age Estimation (CILF-CIAE). Specifically, we first introduce the CLIP model to extract image features and text semantic information respectively, and map them into a highly semantically aligned high-dimensional feature space. Next, we designed a new Transformer architecture (i.e., FourierFormer) to achieve channel evolution and spatial interaction of images, and to fuse image and text semantic information. Compared with the quadratic complexity of the attention mechanism, the proposed Fourierformer is of linear log complexity. To further narrow the semantic gap between image and text features, we utilize an efficient contrastive multimodal learning module that supervises the multimodal fusion process of FourierFormer through contrastive loss for image-text matching, thereby improving the interaction effect between different modalities. Finally, we introduce reversible age estimation, which uses end-to-end error feedback to reduce the error rate of age predictions. Through extensive experiments on multiple data sets, CILF-CIAE has achieved better age prediction results.
Existing automatic evaluation on text-to-image synthesis can only provide an image-text matching score, without considering the object-level compositionality, which results in poor correlation with human judgments. In this work, we propose LLMScore, a new framework that offers evaluation scores with multi-granularity compositionality. LLMScore leverages the large language models (LLMs) to evaluate text-to-image models. Initially, it transforms the image into image-level and object-level visual descriptions. Then an evaluation instruction is fed into the LLMs to measure the alignment between the synthesized image and the text, ultimately generating a score accompanied by a rationale. Our substantial analysis reveals the highest correlation of LLMScore with human judgments on a wide range of datasets (Attribute Binding Contrast, Concept Conjunction, MSCOCO, DrawBench, PaintSkills). Notably, our LLMScore achieves Kendall's tau correlation with human evaluations that is 58.8% and 31.2% higher than the commonly-used text-image matching metrics CLIP and BLIP, respectively.
From image-text pairs, large-scale vision-language models (VLMs) learn to implicitly associate image regions with words, which prove effective for tasks like visual question answering. However, leveraging the learned association for open-vocabulary semantic segmentation remains a challenge. In this paper, we propose a simple, yet extremely effective, training-free technique, Plug-and-Play Open-Vocabulary Semantic Segmentation (PnP-OVSS) for this task. PnP-OVSS leverages a VLM with direct text-to-image cross-attention and an image-text matching loss. To balance between over-segmentation and under-segmentation, we introduce Salience Dropout; by iteratively dropping patches that the model is most attentive to, we are able to better resolve the entire extent of the segmentation mask. PnP-OVSS does not require any neural network training and performs hyperparameter tuning without the need for any segmentation annotations, even for a validation set. PnP-OVSS demonstrates substantial improvements over comparable baselines (+26.2% mIoU on Pascal VOC, +20.5% mIoU on MS COCO, +3.1% mIoU on COCO Stuff and +3.0% mIoU on ADE20K). Our codebase is at https://github.com/letitiabanana/PnP-OVSS.
U.S. Federal Regulators receive over one million comment letters each year from businesses, interest groups, and members of the public, all advocating for changes to proposed regulations. These comments are believed to have wide-ranging impacts on public policy. However, measuring the impact of specific comments is challenging because regulators are required to respond to comments but they do not have to specify which comments they are addressing. In this paper, we propose a simple yet effective solution to this problem by using an iterative contrastive method to train a neural model aiming for matching text from public comments to responses written by regulators. We demonstrate that our proposal substantially outperforms a set of selected text-matching baselines on a human-annotated test set. Furthermore, it delivers performance comparable to the most advanced gigantic language model (i.e., GPT-4), and is more cost-effective when handling comments and regulator responses matching in larger scale.
Advances in multimodal models have greatly improved how interactions relevant to various tasks are modeled. Today's multimodal models mainly focus on the correspondence between images and text, using this for tasks like image-text matching. However, this covers only a subset of real-world interactions. Novel interactions, such as sarcasm expressed through opposing spoken words and gestures or humor expressed through utterances and tone of voice, remain challenging. In this paper, we introduce an approach to enhance multimodal models, which we call Multimodal Mixtures of Experts (MMoE). The key idea in MMoE is to train separate expert models for each type of multimodal interaction, such as redundancy present in both modalities, uniqueness in one modality, or synergy that emerges when both modalities are fused. On a sarcasm detection task (MUStARD) and a humor detection task (URFUNNY), we obtain new state-of-the-art results. MMoE is also able to be applied to various types of models to gain improvement.
Recently, commonsense learning has been a hot topic in image-text matching. Although it can describe more graphic correlations, commonsense learning still has some shortcomings: 1) The existing methods are based on triplet semantic similarity measurement loss, which cannot effectively match the intractable negative in image-text sample pairs. 2) The weak generalization ability of the model leads to the poor effect of image and text matching on large-scale datasets. According to these shortcomings. This paper proposes a novel image-text matching model, called Active Mining Sample Pair Semantics image-text matching model (AMSPS). Compared with the single semantic learning mode of the commonsense learning model with triplet loss function, AMSPS is an active learning idea. Firstly, the proposed Adaptive Hierarchical Reinforcement Loss (AHRL) has diversified learning modes. Its active learning mode enables the model to more focus on the intractable negative samples to enhance the discriminating ability. In addition, AMSPS can also adaptively mine more hidden relevant semantic representations from uncommented items, which greatly improves the performance and generalization ability of the model. Experimental results on Flickr30K and MSCOCO universal datasets show that our proposed method is superior to advanced comparison methods.
Image-text retrieval is a widely studied topic in the field of computer vision due to the exponential growth of multimedia data, whose core concept is to measure the similarity between images and text. However, most existing retrieval methods heavily rely on cross-attention mechanisms for cross-modal fine-grained alignment, which takes into account excessive irrelevant regions and treats prominent and non-significant words equally, thereby limiting retrieval accuracy. This paper aims to investigate an alignment approach that reduces the involvement of non-significant fragments in images and text while enhancing the alignment of prominent segments. For this purpose, we introduce the Cross-Modal Prominent Fragments Enhancement Aligning Network(CPFEAN), which achieves improved retrieval accuracy by diminishing the participation of irrelevant regions during alignment and relatively increasing the alignment similarity of prominent words. Additionally, we incorporate prior textual information into image regions to reduce misalignment occurrences. In practice, we first design a novel intra-modal fragments relationship reasoning method, and subsequently employ our proposed alignment mechanism to compute the similarity between images and text. Extensive quantitative comparative experiments on MS-COCO and Flickr30K datasets demonstrate that our approach outperforms state-of-the-art methods by about 5% to 10% in the rSum metric.
In this paper, we study a challenging task of zero-shot referring image segmentation. This task aims to identify the instance mask that is most related to a referring expression without training on pixel-level annotations. Previous research takes advantage of pre-trained cross-modal models, e.g., CLIP, to align instance-level masks with referring expressions. %Yet, CLIP only considers image-text pair level alignment, which neglects fine-grained image region and complex sentence matching. Yet, CLIP only considers the global-level alignment of image-text pairs, neglecting fine-grained matching between the referring sentence and local image regions. To address this challenge, we introduce a Text Augmented Spatial-aware (TAS) zero-shot referring image segmentation framework that is training-free and robust to various visual encoders. TAS incorporates a mask proposal network for instance-level mask extraction, a text-augmented visual-text matching score for mining the image-text correlation, and a spatial rectifier for mask post-processing. Notably, the text-augmented visual-text matching score leverages a $P$ score and an $N$-score in addition to the typical visual-text matching score. The $P$-score is utilized to close the visual-text domain gap through a surrogate captioning model, where the score is computed between the surrogate model-generated texts and the referring expression. The $N$-score considers the fine-grained alignment of region-text pairs via negative phrase mining, encouraging the masked image to be repelled from the mined distracting phrases. Extensive experiments are conducted on various datasets, including RefCOCO, RefCOCO+, and RefCOCOg. The proposed method clearly outperforms state-of-the-art zero-shot referring image segmentation methods.
Natural scene analysis and remote sensing imagery offer immense potential for advancements in large-scale language-guided context-aware data utilization. This potential is particularly significant for enhancing performance in downstream tasks such as object detection and segmentation with designed language prompting. In light of this, we introduce the CPSeg, Chain-of-Thought Language Prompting for Finer-grained Semantic Segmentation), an innovative framework designed to augment image segmentation performance by integrating a novel "Chain-of-Thought" process that harnesses textual information associated with images. This groundbreaking approach has been applied to a flood disaster scenario. CPSeg encodes prompt texts derived from various sentences to formulate a coherent chain-of-thought. We propose a new vision-language dataset, FloodPrompt, which includes images, semantic masks, and corresponding text information. This not only strengthens the semantic understanding of the scenario but also aids in the key task of semantic segmentation through an interplay of pixel and text matching maps. Our qualitative and quantitative analyses validate the effectiveness of CPSeg.