Given a reference image, composed image retrieval aims to search for the target image from the image gallery with a modification sentence. The complicated semantic relationships within the composed query to comprehend the preserved visual context and modified semantics pose a challenge for the researchers. Most current approaches concentrate on the global or fragmental composition to combine the multi-modal query features and bridge the distance between the query and the target. However, global-level matching tends to ignore fine-grained components that distinguish the matched instance from hard negative instances. As for the fragment-level matching, the over-sensitivity to the local noise due to the lack of explicit local labels and excessive computational demands may have a detrimental impact on performance. In this paper, we propose Semantic Composition via Optimal Transport (SCOT) for composed image retrieval to address the aforementioned issues. To optimize the computational resources, SCOT first introduces Hybrid-modal Affinity Summarization (HAS) to compact salient relationships in the affinity matrix from the reference image patches and modification tokens. Furthermore, to mitigate the over-confidence in local noise, SCOT adaptively assigns optimal transport weights on the salient aligned fragments to Retain Prototype from Alignments (RAS) and guides discriminant semantics learning to Update Semantics from Enhanced Embeddings (USEE), while preserving the characteristics from global structures. Specifically, RAS module investigates the preserved semantics from the reference image based on the affinity matrix, meanwhile USEE module distills significant modification semantics. The proposed SCOT has demonstrated advanced experimental performances on widely-used datasets FashionIQ, CIRR, and Fashion200K. Ablative studies manifest the effectiveness of optimal transport guidance on preserved visual prototypes and updated semantics.

The performance of multimodal models often deteriorates when modality absence occurs. The absence disrupts the learned inter-modal correlations, resulting in biased multimodal representations. This challenge is especially pronounced when the absence is pervasive, affecting both the training and inference phases. Recent studies have attempted to reconstruct the missing information; however, most of them require complete supervision, which is seldom available in scenarios of pervasive absence. The quality of reconstruction remains a critical issue. Alternatively, others aim to learn robust representations from the available modalities but the substantial variations and biases are not fully addressed. This paper introduces the Multimodal Generalization and Refinement (MGR) framework to mitigate the issue of pervasive modality absence. MGR begins by acquiring generalized multimodal representations and iteratively refines them to recognize and calibrate the biased representations. Initially, multimodal samples with absence are embedded through foundation models, and MGR integrates independent unimodal features to further enhance generalization. Additionally, a novel mixed-context prompt is adopted to identify biases in both features and correlations. A redistribution operation can then refine these biases through graph pooling, culminating in robust and calibrated multimodal representations, which are suitable for downstream tasks. Comprehensive experiments on four benchmark datasets demonstrate that the proposed MGR framework outperforms state-of-the-art methods, effectively mitigating the impact of pervasive modality absence.

The task of composed image retrieval aims to match the multi-modal query composed of a reference image and a modification sentence with the target image. Most current approaches narrow the distances between the composed queries and targets by investigating matched correspondences in positive triplets. Nevertheless, they are inclined to exhibit heavy reliance on partial correlations. As the negative correspondences are underestimated, semantic clues that distinguish the target from mismatched candidates are obscured by incomplete associations. Moreover, the correlations between the modification textual features and the visual variations from the reference to candidates are imperative to further strengthen the semantic discriminations. In this paper, we propose DIscriminative Perception from NEgative Correspondences (DIPNEC) to address the aforementioned issues. To encourage awareness of the differences between matched and mismatched correspondences, DIPNEC introduces optimal transport with semantic preservation for reassignments on hard negative triplets. Besides, Difference Quantization Alignments (DQA) and Composed Word-level Alignments (CWA) jointly determine the matching scores between multi-modal queries and candidates. Specifically, DQA concentrates on the correlations of textual features with source-to-target visual differences, and CWA further emphasizes the differentiated semantics. DIPNEC has demonstrated competitive performances on the experimental results and ablation studies on widely-used datasets FashionIQ and CIRR.

Text-to-image (T2I) diffusion models have achieved remarkable progress in generating realistic images from textual descriptions. However, ensuring consistent high-quality image generation with complete backgrounds, object appearance, and optimal texture rendering remains challenging. This paper presents a novel fine-grained pixel-level image editing method based on pre-trained diffusion models. The proposed dual-branch architecture, consisting of Guidance and Generation branches, employs U-Net Denoisers and Self-Attention mechanisms. An improved DDIM-like inversion method obtains the latent representation, followed by multiple denoising steps. Cross-branch interactions, such as KV Replacement, Classifier Guidance, and Feature Correspondence, enable precise control while preserving image fidelity. The iterative refinement and reconstruction process facilitates finegrained editing control, supporting attribute modification, image outpainting, style transfer, and face synthesis with Clickand-Drag style editing using masks. Experimental results demonstrate the effectiveness of the proposed approach in enhancing the quality and controllability of T2I-generated images, surpassing existing methods while maintaining attractive computational complexity for practical real-world applications.

Federated Hetero-Task Learning (FHTL) enables the simultaneous learning of multiple heterogeneous tasks on federated learning clients, offering enhanced flexibility. However, the inconsistency between optimization objectives and evaluation metrics for these heterogeneous tasks poses challenges in achieving performance fairness among clients. This study proposes a fairness-aware FHTL method, FairFHTL. It employs adversarial multi-task representation learning at the client level to learn the task-independent shared model. Consequently, it solves optimization objectives inspired by fair resource allocation on the server side to determine the update direction of the global shared model, ultimately achieving task-independent fair performance balance. Extensive experiments on three multi-task datasets demonstrate that FairFHTL significantly enhances performance across the majority of tasks compared to conventional federated learning and FHTL methods. Moreover, compared with other fairness-aware federated learning approaches, FairFHTL maintains a more uniform performance distribution across all tasks.

The interactive task of composed image retrieval aims to retrieve the most relevant images with the bi-modal query, consisting of a reference image and a modification sentence. Despite significant efforts to bridge the heterogeneous gap within the bi-modal query and leverage contrastive learning to reduce the disparity between positive and negative triplets, prior methods often fail to ensure reliable matching due to aleatoric and epistemic uncertainty. Specifically, the aleatoric uncertainty stems from underlying semantic correlations within candidate instances and annotation noise, and the epistemic uncertainty is usually caused by overconfidence in dominant semantic categories. In this paper, we propose Robust UNcertainty Calibration (RUNC) to quantify the uncertainty and calibrate the imbalanced semantic distribution. To mitigate semantic ambiguity in similarity distribution between fusion queries and targets, RUNC maximizes the matching evidence by utilizing a high-order conjugate prior distribution to fit the semantic covariances in candidate samples. With the estimated uncertainty coefficient of each candidate, the target distribution is calibrated to encourage balanced semantic alignment. Additionally, we minimize the ambiguity in the fusion evidence when forming the unified query by incorporating orthogonal constraints on explicit textual embeddings and implicit queries, to reduce the representation redundancy. Extensive experiments and ablation analysis on benchmark datasets FashionIQ and CIRR verify the robustness of RUNC in predicting reliable retrieval results from a large image gallery.

Automatic generation of ophthalmic reports holds significant potential to lessen clinicians’ workload, enhance work efficiency, and alleviate the imbalance between clinicians and patients. Recent advancements in multimodal large language models, represented by GPT-4, have demonstrated remarkable performance in the general domain. However, training such models necessitates a substantial amount of paired image-text data, yet paired ophthalmic data is limited, and ophthalmic reports are laden with specialized terminologies, making it challenging to transfer the training paradigm to the ophthalmic domain. In this paper, we propose EyeGraphGPT, a knowledge graph enhanced multimodal large language model for ophthalmic report generation. Specifically, we construct a knowledge graph by leveraging the knowledge from a medical database and expertise from ophthalmic experts to model relationships among ophthalmic diseases, enhancing the model’s focus on key disease information. We then perform relation-aware modal alignment to incorporate knowledge graph features into visual features, and further enhance modality collaboration through visual instruction fine-tuning to adapt the model to the ophthalmic domain. Our experiments on a real-world dataset demonstrates that EyeGraphGPT outperforms previous state-of-the-art models, highlighting its superiority in scenarios with limited medical data and extensive specialized terminologies.

Federated learning (FL) enables collaborative learning across multiple biomedical data silos with multimodal foundation models while preserving privacy. Due to the heterogeneity in data processing and collection methodologies across diverse medical institutions and the varying medical inspections patients undergo, modal heterogeneity exists in practical scenarios, where severe modal heterogeneity may even prevent model training. With privacy considerations, data transfer cannot be permitted, restricting knowledge exchange among different clients. To trickle these issues, we propose a cross-modal prototype imputation method for visual-language understanding (Buffalo) with only a slight increase in communication cost, which can improve the performance of fine-tuning general foundation models for downstream biomedical tasks. We conducted extensive experiments on medical report generation and biomedical visual question-answering tasks. The results demonstrate that Buffalo can fully utilize data from all clients to improve model generalization compared to other modal imputation methods in three modal heterogeneity scenarios, approaching or even surpassing the performance in the ideal scenario without missing modality.

Federated Learning has emerged as a revolutionary innovation in the evolving landscape of global healthcare, fostering collaboration among institutions and facilitating collaborative data analysis. As practical applications continue to proliferate, numerous federations have formed in different regions. The optimization and sustainable development of federation-pretrained models have emerged as new challenges. These challenges primarily encompass privacy, population shift and data dependency, which may lead to severe consequences such as the leakage of sensitive information within models and training samples, unfair model performance and resource burdens. To tackle these issues, we propose FairFusion, a cross-federation model fusion approach that enhances privacy and fairness. FairFusion operates across federations within a Model Trip paradigm, integrating knowledge from diverse federations to continually enhance model performance. Through federated model fusion, multi-objective quantification and optimization, FairFusion obtains trustworthy solutions that excel in utility, privacy and fairness. We conduct comprehensive experiments on three public real-world healthcare datasets. The results demonstrate that FairFusion achieves outstanding model fusion performance in terms of utility and fairness across various model structures and subgroups with sensitive attributes while guaranteeing model privacy.

Multi-modality physiological signal-based emotion recognition has attracted increasing attention as its capacity to capture human affective states comprehensively. Due to multi-modality heterogeneity and cross-subject divergence, practical applications struggle with generalizing models across individuals. Effectively addressing both issues requires mitigating the gap between multimodal signals while acquiring generalizable representations across subjects. However, existing approaches often handle these dual challenges separately, resulting in suboptimal generalization. This study introduces a novel framework, termed Correlation-Driven Multi-Modality Graph Decomposition (CMMGD). The proposed CMMGD initially captures adaptive cross-modal correlations. It connects each unimodal graph to a multimodal mixed graph. To simultaneously address the dual challenges, it incorporates a correlation-driven graph decomposition module that decomposes the mixed graph into concordant and discrepant subgraphs based on the correlations. The decomposed concordant subgraph encompasses consistently activated features across modalities and subjects during emotion elicitation, unveiling a generalizable subspace. Additionally, we design a Multi-Modality Graph Regularized Transformer (MGRT) backbone specifically tailored for multimodal physiological signals. The MGRT can alleviate the over-smoothing issue and mitigate over-reliance on any single modality. Extensive experiments demonstrate that CMMGD outperforms the state-of-the-art methods by 1.79% and 2.65% on DEAP and MAHNOB-HCI datasets, respectively, under the leave-one-subject-out cross-validation strategy.

The challenging task composed image retrieval targets at identifying the matched image from the multi-modal query with a reference image and a textual modifier. Most existing methods are devoted to composing the unified query representations from the query images and texts, yet the distribution gaps between the hybrid-modal query representations and visual target representations are neglected. However, directly incorporating target features on the query may cause ambiguous rankings and poor robustness due to the insufficient exploration of the distinguishments and overfitting issues. To address the above concerns, we propose a novel framework termed SemAntic Distillation from Neighborhood (SADN) for composed image retrieval. For mitigating the distribution divergences, we construct neighborhood sampling from the target domain for each query and aggregate neighborhood features with adaptive weights to restructure the query representations. Specifically, the adaptive weights are determined by the collaboration of two individual modules, as correspondence-induced adaption and divergence-based correction. Correspondence-induced adaption accounts for capturing the correlation alignments from neighbor features under the guidance of the positive representations, and the divergence-based correction regulates the weights based on the embedding distances between hard negatives and the query in the latent space. Extensive results and ablation studies on CIRR and FashionIQ validate that the proposed semantic distillation from neighborhood significantly outperforms baseline methods.

Model fusion is becoming a crucial component in the context of model-as-a-service scenarios, enabling the delivery of high-quality model services to local users. However, this approach introduces privacy risks and imposes certain limitations on its applications. Ensuring secure model exchange and knowledge fusion among users becomes a significant challenge in this setting. To tackle this issue, we propose PrivFusion, a novel architecture that preserves privacy while facilitating model fusion under the constraints of local differential privacy. PrivFusion leverages a graph-based structure, enabling the fusion of models from multiple parties without additional training. By employing randomized mechanisms, PrivFusion ensures privacy guarantees throughout the fusion process. To enhance model privacy, our approach incorporates a hybrid local differentially private mechanism and decentralized federated graph matching, effectively protecting both activation values and weights. Additionally, we introduce a perturbation filter adapter to alleviate the impact of randomized noise, thereby recovering the utility of the fused model. Through extensive experiments conducted on diverse image datasets and real-world healthcare applications, we provide empirical evidence showcasing the effectiveness of PrivFusion in maintaining model performance while preserving privacy. Our contributions offer valuable insights and practical solutions for secure and collaborative data analysis within the domain of privacy-preserving model fusion.

Physiological signal based ubiquitous computing has garnered significant attention. However, the heterogeneity among multimodal physiological signals poses a critical challenge to practical applications. To traverse this heterogeneity gap, recent studies have focused on establishing inter-modality correlations. Early works only consider coarse-level correlations between the embeddings of each modality. More recent graph-based approaches incorporate prior knowledge-based correlations, although they may not be entirely accurate. In this article, we propose the Graph Joint Fusion (GJFusion) network, which leverages channel-level inter-modality correlations based on a graph joint to mitigate the heterogeneous gap. Our proposed GJFusion first represents each modality as a graph, with each vertex corresponding to a signal channel, and the edges denoting their functional connectivity. We then join each modality by constructing inter-modality correlations for each salient channel using a sampling-based matching method. Discarded channels are transformed into a virtual vertex through a lightweight pooling operation. Subsequently, the fusion network integrates intra- and inter-modality features, enabling multimodal physiological signal fusion. To validate the effectiveness of our method, we select emotional state recognition as the downstream task and conduct comprehensive experiments on two benchmark datasets. The results demonstrate that our proposed GJFusion network surpasses the latest state-of-the-art methods, achieving relative accuracy improvements of 1.22% and 0.81% on the DEAP and MAHNOB-HCI datasets, respectively. Furthermore, visualization experiments of the salient brain regions reveal the presence of interpretable knowledge within the proposed GJFusion model.

Federated learning has emerged as a powerful technique for training robust models while preserving data privacy and security. However, real-world applications, especially in domains like healthcare, often face challenges due to non-independent and non-identically distributed (non-iid) data across different institutions. Additionally, the heterogeneity of data and the absence of a trusted central server further hinder collaborative efforts among medical institutions. Our paper introduces a novel federated learning approach called FedTAM, which incorporates cyclic model transfer and feature attention-based multi-teacher knowledge distillation. FedTAM is designed to tailor personalized models for individual clients within a decentralized federated learning setting, where data distribution is non-iid. Notably, this method enables student clients to selectively acquire the most pertinent and valuable knowledge from teacher clients through feature attention mechanism while filtering out irrelevant information. We conduct extensive experiments across five benchmark healthcare datasets and one public image classification dataset with feature shifts. Our results conclusively demonstrate that our method achieves remarkable accuracy improvements when compared to state-of-the-art approaches. This affirms the potential of FedTAM to significantly enhance federated learning performance, especially in challenging real-world contexts like healthcare.

Attention Deficit Hyperactivity Disorder (ADHD) is one of the most common mental disorders affecting children. Because the etiology of ADHD is complex and its symptoms are not specific, there is a lack of feasible quantitative diagnostic methods. Pursuing objective and non-invasive detection methods and standards is of great practical significance to prevent the development of the disease. In this study, we aim to address one specific concern about the objectivity and quantification of ADHD diagnosis. Over a year, we iteratively designed and tested WeDA, a scale-driven wearable diagnostic assessment system. This system contains an Android computer machine with a large touchscreen, a suite of 3D printed interactive devices, and six wearable motion sensors. We implement ten diagnostic tasks drawing on the symptoms of ADHD based on DSM-5. The experimental results of classifying children with ADHD and typically developing children and subjective evaluations from doctors, parents, and children validate the effectiveness and acceptability of WeDA.