Emotion recognition is essential across numerous fields, including medical applications and brain-computer interface (BCI). Emotional responses include behavioral reactions, such as tone of voice and body movement, and changes in physiological signals, such as the electroencephalogram (EEG). The latter are involuntary, thus they provide a reliable input for identifying emotions, in contrast to the former which individuals can consciously control. These signals reveal true emotional states without intentional alteration, thus increasing the accuracy of emotion recognition models. However, multimodal deep learning methods from physiological signals have not been significantly investigated. In this paper, we introduce PHemoNet, a fully hypercomplex network for multimodal emotion recognition from physiological signals. In detail, the architecture comprises modality-specific encoders and a fusion module. Both encoders and fusion modules are defined in the hypercomplex domain through parameterized hypercomplex multiplications (PHMs) that can capture latent relations between the different dimensions of each modality and between the modalities themselves. The proposed method outperforms current state-of-the-art models on the MAHNOB-HCI dataset in classifying valence and arousal using electroencephalograms (EEGs) and peripheral physiological signals. The code for this work is available at https://github.com/ispamm/MHyEEG.
Inspired by human visual attention, deep neural networks have widely adopted attention mechanisms to learn locally discriminative attributes for challenging visual classification tasks. However, existing approaches primarily emphasize the representation of such features while neglecting their precise localization, which often leads to misclassification caused by shortcut biases. This limitation becomes even more pronounced when models are evaluated on transfer or out-of-distribution datasets. In contrast, humans are capable of leveraging prior object knowledge to quickly localize and compare fine-grained attributes, a capability that is especially crucial in complex and high-variance classification scenarios. Motivated by this, we introduce Gaze-CIFAR-10, a human gaze time-series dataset, along with a dual-sequence gaze encoder that models the precise sequential localization of human attention on distinct local attributes. In parallel, a Vision Transformer (ViT) is employed to learn the sequential representation of image content. Through cross-modal fusion, our framework integrates human gaze priors with machine-derived visual sequences, effectively correcting inaccurate localization in image feature representations. Extensive qualitative and quantitative experiments demonstrate that gaze-guided cognitive cues significantly enhance classification accuracy.
Molecular subtyping of breast cancer is crucial for personalized treatment and prognosis. Traditional classification approaches rely on either histopathological images or gene expression profiling, limiting their predictive power. In this study, we propose a deep multimodal learning framework that integrates histopathological images and gene expression data to classify breast cancer into BRCA.Luminal and BRCA.Basal / Her2 subtypes. Our approach employs a ResNet-50 model for image feature extraction and fully connected layers for gene expression processing, with a cross-attention fusion mechanism to enhance modality interaction. We conduct extensive experiments using five-fold cross-validation, demonstrating that our multimodal integration outperforms unimodal approaches in terms of classification accuracy, precision-recall AUC, and F1-score. Our findings highlight the potential of deep learning for robust and interpretable breast cancer subtype classification, paving the way for improved clinical decision-making.
Neural networks are one tool for approximating non-linear differential equations used in scientific computing tasks such as surrogate modeling, real-time predictions, and optimal control. PDE foundation models utilize neural networks to train approximations to multiple differential equations simultaneously and are thus a general purpose solver that can be adapted to downstream tasks. Current PDE foundation models focus on either learning general solution operators and/or the governing system of equations, and thus only handle numerical or symbolic modalities. However, real-world applications may require more flexible data modalities, e.g. text analysis or descriptive outputs. To address this gap, we propose a novel multimodal deep learning approach that leverages a transformer-based architecture to approximate solution operators for a wide variety of ODEs and PDEs. Our method integrates numerical inputs, such as equation parameters and initial conditions, with text descriptions of physical processes or system dynamics. This enables our model to handle settings where symbolic representations may be incomplete or unavailable. In addition to providing accurate numerical predictions, our approach generates interpretable scientific text descriptions, offering deeper insights into the underlying dynamics and solution properties. The numerical experiments show that our model provides accurate solutions for in-distribution data (with average relative error less than 3.3%) and out-of-distribution data (average relative error less than 7.8%) together with precise text descriptions (with correct descriptions generated 100% of times). In certain tests, the model is also shown to be capable of extrapolating solutions in time.
This study presents a comprehensive review of the potential of multimodal deep learning (DL) in medical diagnosis, using COVID-19 as a case example. Motivated by the success of artificial intelligence applications during the COVID-19 pandemic, this research aims to uncover the capabilities of DL in disease screening, prediction, and classification, and to derive insights that enhance the resilience, sustainability, and inclusiveness of science, technology, and innovation systems. Adopting a systematic approach, we investigate the fundamental methodologies, data sources, preprocessing steps, and challenges encountered in various studies and implementations. We explore the architecture of deep learning models, emphasising their data-specific structures and underlying algorithms. Subsequently, we compare different deep learning strategies utilised in COVID-19 analysis, evaluating them based on methodology, data, performance, and prerequisites for future research. By examining diverse data types and diagnostic modalities, this research contributes to scientific understanding and knowledge of the multimodal application of DL and its effectiveness in diagnosis. We have implemented and analysed 11 deep learning models using COVID-19 image, text, and speech (ie, cough) data. Our analysis revealed that the MobileNet model achieved the highest accuracy of 99.97% for COVID-19 image data and 93.73% for speech data (i.e., cough). However, the BiGRU model demonstrated superior performance in COVID-19 text classification with an accuracy of 99.89%. The broader implications of this research suggest potential benefits for other domains and disciplines that could leverage deep learning techniques for image, text, and speech analysis.
This study introduces CLASP (Contrastive Language-Speech Pretraining), a multilingual, multimodal representation tailored for audio-text information retrieval. CLASP leverages the synergy between spoken content and textual data. During training, we utilize our newly introduced speech-text dataset, which encompasses 15 diverse categories ranging from fiction to religion. CLASP's audio component integrates audio spectrograms with a pre-trained self-supervised speech model, while its language encoding counterpart employs a sentence encoder pre-trained on over 100 languages. This unified lightweight model bridges the gap between various modalities and languages, enhancing its effectiveness in handling and retrieving multilingual and multimodal data. Our evaluations across multiple languages demonstrate that CLASP establishes new benchmarks in HITS@1, MRR, and meanR metrics, outperforming traditional ASR-based retrieval methods that rely on transcribing speech into text for subsequent text retrieval, especially in specific scenarios.
Background: Laboratory values are fundamental to medical diagnosis and management, but acquiring these values can be costly, invasive, and time-consuming. While electrocardiogram (ECG) patterns have been linked to certain laboratory abnormalities, the comprehensive modeling of these relationships remains underexplored. Methods: We utilize MIMIC-IV dataset to develop multimodal deep-learning models to demonstrate the feasibility of estimating (real-time) and monitoring (predict at future intervals) laboratory value abnormalities from ECG waveforms, demographics, biometrics, and vital signs. Results: The models exhibit a strong predictive performance with AUROC scores above 0.70 in a statistically significant manner for 23 laboratory values in the estimation setting and up to 26 values in the monitoring setting. Most notably, the accurately predictable values encompassing abnormalities across diverse physiological categories such as cardiac, renal, hematological, metabolic, immunological and coagulation. To name examples, for estimation NTproBNP (>353 pg/mL) with 0.882, whereas for monitoring at 30 minutes Urea nitrogen (<6 mg/dL) with 0.851, at 60 minutes creatinine (<0.5 mg/dL) with 0.85, and at 120 minutes hemoglobin (>17.5 g/dL) with 0.821. Conclusions: This study provides first evidence for the feasibility of using ECG data alongside clinical routine data for the real-time estimation and monitoring of laboratory value abnormalities, which could provide a non-invasive, cost-effective supplement to traditional laboratory testing, with strong implications for enhanced patient monitoring and early intervention. Further validation could facilitate their integration into routine clinical practice.
Glioblastoma is a highly aggressive form of brain cancer characterized by rapid progression and poor prognosis. Despite advances in treatment, the underlying genetic mechanisms driving this aggressiveness remain poorly understood. In this study, we employed multimodal deep learning approaches to investigate glioblastoma heterogeneity using joint image/RNA-seq analysis. Our results reveal novel genes associated with glioblastoma. By leveraging a combination of whole-slide images and RNA-seq, as well as introducing novel methods to encode RNA-seq data, we identified specific genetic profiles that may explain different patterns of glioblastoma progression. These findings provide new insights into the genetic mechanisms underlying glioblastoma heterogeneity and highlight potential targets for therapeutic intervention. Code and data downloading instructions are available at: https://github.com/ma3oun/gbheterogeneity.
Image clustering, which involves grouping images into different clusters without labels, is a key task in unsupervised learning. Although previous deep clustering methods have achieved remarkable results, they only explore the intrinsic information of the image itself but overlook external supervision knowledge to improve the semantic understanding of images. Recently, visual-language pre-trained model on large-scale datasets have been used in various downstream tasks and have achieved great results. However, there is a gap between visual representation learning and textual semantic learning, and how to properly utilize the representation of two different modalities for clustering is still a big challenge. To tackle the challenges, we propose a novel image clustering framwork, named Dual-level Cross-Modal Contrastive Clustering (DXMC). Firstly, external textual information is introduced for constructing a semantic space which is adopted to generate image-text pairs. Secondly, the image-text pairs are respectively sent to pre-trained image and text encoder to obtain image and text embeddings which subsquently are fed into four well-designed networks. Thirdly, dual-level cross-modal contrastive learning is conducted between discriminative representations of different modalities and distinct level. Extensive experimental results on five benchmark datasets demonstrate the superiority of our proposed method.
Recently, multimodal deep learning, which integrates histopathology slides and molecular biomarkers, has achieved a promising performance in glioma grading. Despite great progress, due to the intra-modality complexity and inter-modality heterogeneity, existing studies suffer from inadequate histopathology representation learning and inefficient molecular-pathology knowledge alignment. These two issues hinder existing methods to precisely interpret diagnostic molecular-pathology features, thereby limiting their grading performance. Moreover, the real-world applicability of existing multimodal approaches is significantly restricted as molecular biomarkers are not always available during clinical deployment. To address these problems, we introduce a novel Focus on Focus (FoF) framework with paired pathology-genomic training and applicable pathology-only inference, enhancing molecular-pathology representation effectively. Specifically, we propose a Focus-oriented Representation Learning (FRL) module to encourage the model to identify regions positively or negatively related to glioma grading and guide it to focus on the diagnostic areas with a consistency constraint. To effectively link the molecular biomarkers to morphological features, we propose a Multi-view Cross-modal Alignment (MCA) module that projects histopathology representations into molecular subspaces, aligning morphological features with corresponding molecular biomarker status by supervised contrastive learning. Experiments on the TCGA GBM-LGG dataset demonstrate that our FoF framework significantly improves the glioma grading. Remarkably, our FoF achieves superior performance using only histopathology slides compared to existing multimodal methods. The source code is available at https://github.com/peterlipan/FoF.
Multimodal Deep Learning enhances decision-making by integrating diverse information sources, such as texts, images, audio, and videos. To develop trustworthy multimodal approaches, it is essential to understand how uncertainty impacts these models. We propose LUMA, a unique benchmark dataset, featuring audio, image, and textual data from 50 classes, for learning from uncertain and multimodal data. It extends the well-known CIFAR 10/100 dataset with audio samples extracted from three audio corpora, and text data generated using the Gemma-7B Large Language Model (LLM). The LUMA dataset enables the controlled injection of varying types and degrees of uncertainty to achieve and tailor specific experiments and benchmarking initiatives. LUMA is also available as a Python package including the functions for generating multiple variants of the dataset with controlling the diversity of the data, the amount of noise for each modality, and adding out-of-distribution samples. A baseline pre-trained model is also provided alongside three uncertainty quantification methods: Monte-Carlo Dropout, Deep Ensemble, and Reliable Conflictive Multi-View Learning. This comprehensive dataset and its benchmarking tools are intended to promote and support the development, evaluation, and benchmarking of trustworthy and robust multimodal deep learning approaches. We anticipate that the LUMA dataset will help the ICLR community to design more trustworthy and robust machine learning approaches for safety critical applications.
Plant classification is vital for ecological conservation and agricultural productivity, enhancing our understanding of plant growth dynamics and aiding species preservation. The advent of deep learning (DL) techniques has revolutionized this field by enabling autonomous feature extraction, significantly reducing the dependence on manual expertise. However, conventional DL models often rely solely on single data sources, failing to capture the full biological diversity of plant species comprehensively. Recent research has turned to multimodal learning to overcome this limitation by integrating multiple data types, which enriches the representation of plant characteristics. This shift introduces the challenge of determining the optimal point for modality fusion. In this paper, we introduce a pioneering multimodal DL-based approach for plant classification with automatic modality fusion. Utilizing the multimodal fusion architecture search, our method integrates images from multiple plant organs -- flowers, leaves, fruits, and stems -- into a cohesive model. To address the lack of multimodal datasets, we contributed Multimodal-PlantCLEF, a restructured version of the PlantCLEF2015 dataset tailored for multimodal tasks. Our method achieves 82.61% accuracy on 979 classes of Multimodal-PlantCLEF, surpassing state-of-the-art methods and outperforming late fusion by 10.33%. Through the incorporation of multimodal dropout, our approach demonstrates strong robustness to missing modalities. We validate our model against established benchmarks using standard performance metrics and McNemar's test, further underscoring its superiority.
Developing accurate machine learning models for oncology requires large-scale, high-quality multimodal datasets. However, creating such datasets remains challenging due to the complexity and heterogeneity of medical data. To address this challenge, we introduce HoneyBee, a scalable modular framework for building multimodal oncology datasets that leverages foundation models to generate representative embeddings. HoneyBee integrates various data modalities, including clinical diagnostic and pathology imaging data, medical notes, reports, records, and molecular data. It employs data preprocessing techniques and foundation models to generate embeddings that capture the essential features and relationships within the raw medical data. The generated embeddings are stored in a structured format using Hugging Face datasets and PyTorch dataloaders for accessibility. Vector databases enable efficient querying and retrieval for machine learning applications. We demonstrate the effectiveness of HoneyBee through experiments assessing the quality and representativeness of these embeddings. The framework is designed to be extensible to other medical domains and aims to accelerate oncology research by providing high-quality, machine learning-ready datasets. HoneyBee is an ongoing open-source effort, and the code, datasets, and models are available at the project repository.
Surgery to treat elderly hip fracture patients may cause complications that can lead to early mortality. An early warning system for complications could provoke clinicians to monitor high-risk patients more carefully and address potential complications early, or inform the patient. In this work, we develop a multimodal deep-learning model for post-operative mortality prediction using pre-operative and per-operative data from elderly hip fracture patients. Specifically, we include static patient data, hip and chest images before surgery in pre-operative data, vital signals, and medications administered during surgery in per-operative data. We extract features from image modalities using ResNet and from vital signals using LSTM. Explainable model outcomes are essential for clinical applicability, therefore we compute Shapley values to explain the predictions of our multimodal black box model. We find that i) Shapley values can be used to estimate the relative contribution of each modality both locally and globally, and ii) a modified version of the chain rule can be used to propagate Shapley values through a sequence of models supporting interpretable local explanations. Our findings imply that a multimodal combination of black box models can be explained by propagating Shapley values through the model sequence.
Visual Question Answerinng (VQA) is a complicated task that requires the capability of simultaneously processing natural language and images. This task was initially researched with a focus on developing methods to help machines understand objects and scene contexts in images. However, some scene text that carries explicit information about the full content of the image is not mentioned. Along with the continuous development of the AI era, there have been many studies on the reading comprehension ability of VQA models in the world. Therefore, we introduce the first large-scale dataset in Vietnamese specializing in the ability to understand scene text, we call it ViTextVQA (\textbf{Vi}etnamese \textbf{Text}-based \textbf{V}isual \textbf{Q}uestion \textbf{A}nswering dataset) which contains \textbf{over 16,000} images and \textbf{over 50,000} questions with answers. To tackle this task efficiently, we propose ViTextBLIP-2, an novel multimodal feature fusion Method, which optimizes Vietnamese OCR-based VQA by integrating a frozen Vision Transformer, SwinTextSpotter OCR, and ViT5 LLM with a trainable Q-Former for multimodal feature fusion. Through experiments with various state-of-the-art models, we uncover the significance of the order in which tokens in OCR text are processed and selected to formulate answers. This finding helped us significantly improve the performance of the baseline models on the ViTextVQA dataset. Our dataset is available (https://github.com/minhquan6203/ViTextVQA-Dataset) for research purposes.
Studying and analyzing cropland is a difficult task due to its dynamic and heterogeneous growth behavior. Usually, diverse data sources can be collected for its estimation. Although deep learning models have proven to excel in the crop classification task, they face substantial challenges when dealing with multiple inputs, named Multi-View Learning (MVL). The methods used in the MVL scenario can be structured based on the encoder architecture, the fusion strategy, and the optimization technique. The literature has primarily focused on using specific encoder architectures for local regions, lacking a deeper exploration of other components in the MVL methodology. In contrast, we investigate the simultaneous selection of the fusion strategy and encoder architecture, assessing global-scale cropland and crop-type classifications. We use a range of five fusion strategies (Input, Feature, Decision, Ensemble, Hybrid) and five temporal encoders (LSTM, GRU, TempCNN, TAE, L-TAE) as possible configurations in the MVL method. We use the CropHarvest dataset for validation, which provides optical, radar, weather time series, and topographic information as input data. We found that in scenarios with a limited number of labeled samples, a unique configuration is insufficient for all the cases. Instead, a specialized combination should be meticulously sought, including an encoder and fusion strategy. To streamline this search process, we suggest identifying the optimal encoder architecture tailored for a particular fusion strategy, and then determining the most suitable fusion strategy for the classification task. We provide a methodological framework for researchers exploring crop classification through an MVL methodology.
Despite the demonstrated parameter efficiency of prompt-based multimodal fusion methods, their limited adaptivity and expressiveness often result in suboptimal performance compared to other tuning approaches. In this paper, we introduce the Mixture of Prompt Experts (MoPE), the first technique designed to overcome these limitations by decomposing standard prompts to capture instance-level features adaptively. Building on this decomposition, MoPE enhances prompt fusion's expressiveness by leveraging multimodal pairing priors to route the most effective prompt for each instance dynamically. Compared to vanilla prompting, our MoPE-based fusion method exhibits greater expressiveness, scaling more effectively with the training data and the overall number of trainable parameters. We also investigate regularization terms for expert routing, which lead to emergent expert specialization with enhanced adaptiveness and interpretablity. Extensive experiments across six multimodal datasets spanning four modalities demonstrate state-of-the-art performance for prompt fusion, matching or even surpassing the performance of fine-tuning while requiring only 0.8% of the trainable parameters. Project homepage: https://github.com/songrise/MoPE
Cultural heritage serves as the enduring record of human thought and history. Despite significant efforts dedicated to the preservation of cultural relics, many ancient artefacts have been ravaged irreversibly by natural deterioration and human actions. Deep learning technology has emerged as a valuable tool for restoring various kinds of cultural heritages, including ancient text restoration. Previous research has approached ancient text restoration from either visual or textual perspectives, often overlooking the potential of synergizing multimodal information. This paper proposes a novel Multimodal Multitask Restoring Model (MMRM) to restore ancient texts, particularly emphasising the ideograph. This model combines context understanding with residual visual information from damaged ancient artefacts, enabling it to predict damaged characters and generate restored images simultaneously. We tested the MMRM model through experiments conducted on both simulated datasets and authentic ancient inscriptions. The results show that the proposed method gives insightful restoration suggestions in both simulation experiments and real-world scenarios. To the best of our knowledge, this work represents the pioneering application of multimodal deep learning in ancient text restoration, which will contribute to the understanding of ancient society and culture in digital humanities fields.
We present DeepSeek-VL, an open-source Vision-Language (VL) Model designed for real-world vision and language understanding applications. Our approach is structured around three key dimensions: We strive to ensure our data is diverse, scalable, and extensively covers real-world scenarios including web screenshots, PDFs, OCR, charts, and knowledge-based content, aiming for a comprehensive representation of practical contexts. Further, we create a use case taxonomy from real user scenarios and construct an instruction tuning dataset accordingly. The fine-tuning with this dataset substantially improves the model's user experience in practical applications. Considering efficiency and the demands of most real-world scenarios, DeepSeek-VL incorporates a hybrid vision encoder that efficiently processes high-resolution images (1024 x 1024), while maintaining a relatively low computational overhead. This design choice ensures the model's ability to capture critical semantic and detailed information across various visual tasks. We posit that a proficient Vision-Language Model should, foremost, possess strong language abilities. To ensure the preservation of LLM capabilities during pretraining, we investigate an effective VL pretraining strategy by integrating LLM training from the beginning and carefully managing the competitive dynamics observed between vision and language modalities. The DeepSeek-VL family (both 1.3B and 7B models) showcases superior user experiences as a vision-language chatbot in real-world applications, achieving state-of-the-art or competitive performance across a wide range of visual-language benchmarks at the same model size while maintaining robust performance on language-centric benchmarks. We have made both 1.3B and 7B models publicly accessible to foster innovations based on this foundation model.
Multi-modal foundation models like OpenFlamingo, LLaVA, and GPT-4 are increasingly used for various real-world tasks. Prior work has shown that these models are highly vulnerable to adversarial attacks on the vision modality. These attacks can be leveraged to spread fake information or defraud users, and thus pose a significant risk, which makes the robustness of large multi-modal foundation models a pressing problem. The CLIP model, or one of its variants, is used as a frozen vision encoder in many large vision-language models (LVLMs), e.g. LLaVA and OpenFlamingo. We propose an unsupervised adversarial fine-tuning scheme to obtain a robust CLIP vision encoder, which yields robustness on all vision down-stream tasks (LVLMs, zero-shot classification) that rely on CLIP. In particular, we show that stealth-attacks on users of LVLMs by a malicious third party providing manipulated images are no longer possible once one replaces the original CLIP model with our robust one. No retraining or fine-tuning of the down-stream LVLMs is required. The code and robust models are available at https://github.com/chs20/RobustVLM