The concept of waterbody style transfer remains largely unexplored in the underwater imaging and vision literature. Traditional image style transfer (STx) methods primarily focus on artistic and photorealistic blending, often failing to preserve object and scene geometry in images captured in high-scattering mediums such as underwater. The wavelength-dependent nonlinear attenuation and depth-dependent backscattering artifacts further complicate learning underwater image STx from unpaired data. This paper introduces UStyle, the first data-driven learning framework for transferring waterbody styles across underwater images without requiring prior reference images or scene information. We propose a novel depth-aware whitening and coloring transform (DA-WCT) mechanism that integrates physics-based waterbody synthesis to ensure perceptually consistent stylization while preserving scene structure. To enhance style transfer quality, we incorporate carefully designed loss functions that guide UStyle to maintain colorfulness, lightness, structural integrity, and frequency-domain characteristics, as well as high-level content in VGG and CLIP (contrastive language-image pretraining) feature spaces. By addressing domain-specific challenges, UStyle provides a robust framework for no-reference underwater image STx, surpassing state-of-the-art (SOTA) methods that rely solely on end-to-end reconstruction loss. Furthermore, we introduce the UF7D dataset, a curated collection of high-resolution underwater images spanning seven distinct waterbody styles, establishing a benchmark to support future research in underwater image STx. The UStyle inference pipeline and UF7D dataset are released at: https://github.com/uf-robopi/UStyle.
The histopathological classification of whole-slide images (WSIs) is a fundamental task in digital pathology; yet it requires extensive time and expertise from specialists. While deep learning methods show promising results, they typically process WSIs by dividing them into artificial patches, which inherently prevents a network from learning from the entire image context, disregards natural tissue structures and compromises interpretability. Our method overcomes this limitation through a novel graph-based framework that constructs WSI graph representations. The WSI-graph efficiently captures essential histopathological information in a compact form. We build tissue representations (nodes) that follow biological boundaries rather than arbitrary patches all while providing interpretable features for explainability. Through adaptive graph coarsening guided by learned embeddings, we progressively merge regions while maintaining discriminative local features and enabling efficient global information exchange. In our method's final step, we solve the diagnostic task through a graph attention network. We empirically demonstrate strong performance on multiple challenging tasks such as cancer stage classification and survival prediction, while also identifying predictive factors using Integrated Gradients. Our implementation is publicly available at https://github.com/HistoGraph31/pix2pathology
In clinical imaging, magnetic resonance (MR) image volumes are often acquired as stacks of 2D slices with decreased scan times, improved signal-to-noise ratio, and image contrasts unique to 2D MR pulse sequences. While this is sufficient for clinical evaluation, automated algorithms designed for 3D analysis perform poorly on multi-slice 2D MR volumes, especially those with thick slices and gaps between slices. Super-resolution (SR) methods aim to address this problem, but previous methods do not address all of the following: slice profile shape estimation, slice gap, domain shift, and non-integer or arbitrary upsampling factors. In this paper, we propose ECLARE (Efficient Cross-planar Learning for Anisotropic Resolution Enhancement), a self-SR method that addresses each of these factors. ECLARE uses a slice profile estimated from the multi-slice 2D MR volume, trains a network to learn the mapping from low-resolution to high-resolution in-plane patches from the same volume, and performs SR with anti-aliasing. We compared ECLARE to cubic B-spline interpolation, SMORE, and other contemporary SR methods. We used realistic and representative simulations so that quantitative performance against ground truth can be computed, and ECLARE outperformed all other methods in both signal recovery and downstream tasks. Importantly, as ECLARE does not use external training data it cannot suffer from domain shift between training and testing. Our code is open-source and available at https://www.github.com/sremedios/eclare.
Despite the growing scale of medical Vision-Language datasets, the impact of dataset quality on model performance remains under-explored. We introduce Open-PMC, a high-quality medical dataset from PubMed Central, containing 2.2 million image-text pairs, enriched with image modality annotations, subfigures, and summarized in-text references. Notably, the in-text references provide richer medical context, extending beyond the abstract information typically found in captions. Through extensive experiments, we benchmark Open-PMC against larger datasets across retrieval and zero-shot classification tasks. Our results show that dataset quality-not just size-drives significant performance gains. We complement our benchmark with an in-depth analysis of feature representation. Our findings highlight the crucial role of data curation quality in advancing multimodal medical AI. We release Open-PMC, along with the trained models and our codebase.
Medical image segmentation aims to identify anatomical structures at the voxel-level. Segmentation accuracy relies on distinguishing voxel differences. Compared to advancements achieved in studies of the inter-class variance, the intra-class variance receives less attention. Moreover, traditional linear classifiers, limited by a single learnable weight per class, struggle to capture this finer distinction. To address the above challenges, we propose a Multi-Prototype-based Embedding Refinement method for semi-supervised medical image segmentation. Specifically, we design a multi-prototype-based classification strategy, rethinking the segmentation from the perspective of structural relationships between voxel embeddings. The intra-class variations are explored by clustering voxels along the distribution of multiple prototypes in each class. Next, we introduce a consistency constraint to alleviate the limitation of linear classifiers. This constraint integrates different classification granularities from a linear classifier and the proposed prototype-based classifier. In the thorough evaluation on two popular benchmarks, our method achieves superior performance compared with state-of-the-art methods. Code is available at https://github.com/Briley-byl123/MPER.
Recent advances in Latent Video Diffusion Models (LVDMs) have revolutionized video generation by leveraging Video Variational Autoencoders (Video VAEs) to compress intricate video data into a compact latent space. However, as LVDM training scales, the computational overhead of Video VAEs becomes a critical bottleneck, particularly for encoding high-resolution videos. To address this, we propose LeanVAE, a novel and ultra-efficient Video VAE framework that introduces two key innovations: (1) a lightweight architecture based on a Neighborhood-Aware Feedforward (NAF) module and non-overlapping patch operations, drastically reducing computational cost, and (2) the integration of wavelet transforms and compressed sensing techniques to enhance reconstruction quality. Extensive experiments validate LeanVAE's superiority in video reconstruction and generation, particularly in enhancing efficiency over existing Video VAEs. Our model offers up to 50x fewer FLOPs and 44x faster inference speed while maintaining competitive reconstruction quality, providing insights for scalable, efficient video generation. Our models and code are available at https://github.com/westlake-repl/LeanVAE
Facing the escalating threat of global wildfires, numerous computer vision techniques using remote sensing data have been applied in this area. However, the selection of deep learning methods for wildfire prediction remains uncertain due to the lack of comparative analysis in a quantitative and explainable manner, crucial for improving prevention measures and refining models. This study aims to thoroughly compare the performance, efficiency, and explainability of four prevalent deep learning architectures: Autoencoder, ResNet, UNet, and Transformer-based Swin-UNet. Employing a real-world dataset that includes nearly a decade of remote sensing data from California, U.S., these models predict the spread of wildfires for the following day. Through detailed quantitative comparison analysis, we discovered that Transformer-based Swin-UNet and UNet generally outperform Autoencoder and ResNet, particularly due to the advanced attention mechanisms in Transformer-based Swin-UNet and the efficient use of skip connections in both UNet and Transformer-based Swin-UNet, which contribute to superior predictive accuracy and model interpretability. Then we applied XAI techniques on all four models, this not only enhances the clarity and trustworthiness of models but also promotes focused improvements in wildfire prediction capabilities. The XAI analysis reveals that UNet and Transformer-based Swin-UNet are able to focus on critical features such as 'Previous Fire Mask', 'Drought', and 'Vegetation' more effectively than the other two models, while also maintaining balanced attention to the remaining features, leading to their superior performance. The insights from our thorough comparative analysis offer substantial implications for future model design and also provide guidance for model selection in different scenarios.
Machine-learning interatomic potentials (MLIPs) have greatly extended the reach of atomic-scale simulations, offering the accuracy of first-principles calculations at a fraction of the effort. Leveraging large quantum mechanical databases and expressive architectures, recent "universal" models deliver qualitative accuracy across the periodic table but are often biased toward low-energy configurations. We introduce PET-MAD, a generally applicable MLIP trained on a dataset combining stable inorganic and organic solids, systematically modified to enhance atomic diversity. Using a moderate but highly-consistent level of electronic-structure theory, we assess PET-MAD's accuracy on established benchmarks and advanced simulations of six materials. PET-MAD rivals state-of-the-art MLIPs for inorganic solids, while also being reliable for molecules, organic materials, and surfaces. It is stable and fast, enabling, out-of-the-box, the near-quantitative study of thermal and quantum mechanical fluctuations, functional properties, and phase transitions. It can be efficiently fine-tuned to deliver full quantum mechanical accuracy with a minimal number of targeted calculations.
Cholera, an acute diarrheal disease, is a serious concern in developing and underdeveloped areas. A qualitative understanding of cholera epidemics aims to foresee transmission patterns based on reported data and mechanistic models. The mechanistic model is a crucial tool for capturing the dynamics of disease transmission and population spread. However, using real-time cholera cases is essential for forecasting the transmission trend. This prospective study seeks to furnish insights into transmission trends through qualitative dynamics followed by machine learning-based forecasting. The Monte Carlo Markov Chain approach is employed to calibrate the proposed mechanistic model. We identify critical parameters that illustrate the disease's dynamics using partial rank correlation coefficient-based sensitivity analysis. The basic reproduction number as a crucial threshold measures asymptotic dynamics. Furthermore, forward bifurcation directs the stability of the infection state, and Hopf bifurcation suggests that trends in transmission may become unpredictable as societal disinfection rates rise. Further, we develop epidemic-informed machine learning models by incorporating mechanistic cholera dynamics into autoregressive integrated moving averages and autoregressive neural networks. We forecast short-term future cholera cases in Malawi by implementing the proposed epidemic-informed machine learning models to support this. We assert that integrating temporal dynamics into the machine learning models can enhance the capabilities of cholera forecasting models. The execution of this mechanism can significantly influence future trends in cholera transmission. This evolving approach can also be beneficial for policymakers to interpret and respond to potential disease systems. Moreover, our methodology is replicable and adaptable, encouraging future research on disease dynamics.
Medical ultrasound imaging is ubiquitous, but manual analysis struggles to keep pace. Automated segmentation can help but requires large labeled datasets, which are scarce. Semi-supervised learning leveraging both unlabeled and limited labeled data is a promising approach. State-of-the-art methods use consistency regularization or pseudo-labeling but grow increasingly complex. Without sufficient labels, these models often latch onto artifacts or allow anatomically implausible segmentations. In this paper, we present a simple yet effective pseudo-labeling method with an adversarially learned shape prior to regularize segmentations. Specifically, we devise an encoder-twin-decoder network where the shape prior acts as an implicit shape model, penalizing anatomically implausible but not ground-truth-deviating predictions. Without bells and whistles, our simple approach achieves state-of-the-art performance on two benchmarks under different partition protocols. We provide a strong baseline for future semi-supervised medical image segmentation. Code is available at https://github.com/WUTCM-Lab/Shape-Prior-Semi-Seg.
The carbonic anhydrase II enzyme (CA II) is one of the most significant enzymes in nature, reversibly converting CO$_2$ to bicarbonate at a remarkable rate. The precise mechanism it uses to achieve this rapid turnover remains unclear due to our inability to directly observe or simulate the full process dynamically. Here, we use a recently developed universal neural network potential (Orb) to simulate the active site of CA II. We reproduce several known features of the reaction mechanism, including the proton relay that conducts protons out of the active site to the His64 residue. Additionally, we observe a new reaction pathway where CO$_2$ reacts with a water molecule in the active site, which donates a proton to the zinc-bound hydroxide. This differs from the established mechanism where CO$_2$ directly reacts with hydroxide. Existing experimental data and independent quantum chemistry calculations are used to support the plausibility of this new mechanism. This demonstrates the potential of Orb to efficiently generate novel insights into important molecular scale processes that can potentially be harnessed to improve CO$_2$ capture technologies and drug design.
T2 hyperintensities in spinal cord MR images are crucial biomarkers for conditions such as degenerative cervical myelopathy. However, current clinical diagnoses primarily rely on manual evaluation. Deep learning methods have shown promise in lesion detection, but most supervised approaches are heavily dependent on large, annotated datasets. Unsupervised anomaly detection (UAD) offers a compelling alternative by eliminating the need for abnormal data annotations. However, existing UAD methods rely on curated normal datasets and their performance frequently deteriorates when applied to clinical datasets due to domain shifts. We propose an Uncertainty-based Unsupervised Anomaly Detection framework, termed U2AD, to address these limitations. Unlike traditional methods, U2AD is designed to be trained and tested within the same clinical dataset, following a "mask-and-reconstruction" paradigm built on a Vision Transformer-based architecture. We introduce an uncertainty-guided masking strategy to resolve task conflicts between normal reconstruction and anomaly detection to achieve an optimal balance. Specifically, we employ a Monte-Carlo sampling technique to estimate reconstruction uncertainty mappings during training. By iteratively optimizing reconstruction training under the guidance of both epistemic and aleatoric uncertainty, U2AD reduces overall reconstruction variance while emphasizing regions. Experimental results demonstrate that U2AD outperforms existing supervised and unsupervised methods in patient-level identification and segment-level localization tasks. This framework establishes a new benchmark for incorporating uncertainty guidance into UAD, highlighting its clinical utility in addressing domain shifts and task conflicts in medical image anomaly detection. Our code is available: https://github.com/zhibaishouheilab/U2AD
To this day, accurately simulating local-scale precipitation and reliably reproducing its distribution remains a challenging task. The limited horizontal resolution of Global Climate Models is among the primary factors undermining their skill in this context. The physical mechanisms driving the onset and development of precipitation, especially in extreme events, operate at spatio-temporal scales smaller than those numerically resolved, thus struggling to be captured accurately. In order to circumvent this limitation, several downscaling approaches have been developed over the last decades to address the discrepancy between the spatial resolution of models output and the resolution required by local-scale applications. In this paper, we introduce RainScaleGAN, a conditional deep convolutional Generative Adversarial Network (GAN) for precipitation downscaling. GANs have been effectively used in image super-resolution, an approach highly relevant for downscaling tasks. RainScaleGAN's capabilities are tested in a perfect-model setup, where the spatial resolution of a precipitation dataset is artificially degraded from 0.25$^{\circ}\times$0.25$^{\circ}$ to 2$^{\circ}\times$2$^\circ$, and RainScaleGAN is used to restore it. The developed model outperforms one of the leading precipitation downscaling method found in the literature. RainScaleGAN not only generates a synthetic dataset featuring plausible high-resolution spatial patterns and intensities, but also produces a precipitation distribution with statistics closely mirroring those of the ground-truth dataset. Given that RainScaleGAN's approach is agnostic with respect to the underlying physics, the method has the potential to be applied to other physical variables such as surface winds or temperature.
The prevailing method for neural speech enhancement predominantly utilizes fully-supervised deep learning with simulated pairs of far-field noisy-reverberant speech and clean speech. Nonetheless, these models frequently demonstrate restricted generalizability to mixtures recorded in real-world conditions. To address this issue, this study investigates training enhancement models directly on real mixtures. Specifically, we revisit the single-channel far-field to near-field speech enhancement (FNSE) task, focusing on real-world data characterized by low signal-to-noise ratio (SNR), high reverberation, and mid-to-high frequency attenuation. We propose FNSE-SBGAN, a framework that integrates a Schrodinger Bridge (SB)-based diffusion model with generative adversarial networks (GANs). Our approach achieves state-of-the-art performance across various metrics and subjective evaluations, significantly reducing the character error rate (CER) by up to 14.58% compared to far-field signals. Experimental results demonstrate that FNSE-SBGAN preserves superior subjective quality and establishes a new benchmark for real-world far-field speech enhancement. Additionally, we introduce an evaluation framework leveraging matrix rank analysis in the time-frequency domain, providing systematic insights into model performance and revealing the strengths and weaknesses of different generative methods.
We propose a function-on-function linear regression model for time-dependent curve data that is consistently estimated by imposing factor structures on the regressors. An integral operator based on cross-covariances identifies two components for each functional regressor: a predictive low-dimensional component, along with associated factors that are guaranteed to be correlated with the dependent variable, and an infinite-dimensional component that has no predictive power. In order to consistently estimate the correct number of factors for each regressor, we introduce a functional eigenvalue difference test. Our setting allows us to construct a novel central limit theorem for the regression parameters in a fully functional model, making it possible to construct confidence bands and conduct statistical inference. The model is applied to forecast electricity price curves in three different energy markets. Its prediction accuracy is found to be comparable to popular machine learning approaches, while providing statistically valid inference and interpretable insights into the conditional correlation structures of electricity prices.
Generative machine learning models for small molecule drug discovery have shown immense promise, but many molecules they generate are too difficult to synthesize, making them impractical for further investigation or development. In this work, we present a novel approach by fine-tuning Meta's Llama3 Large Language Models (LLMs) to create SynLlama, which generates full synthetic pathways made of commonly accessible building blocks and robust organic reaction templates. SynLlama explores a large synthesizable space using significantly less data compared to other state-of-the-art methods, and offers strong performance in bottom-up synthesis, synthesizable analog generation, and hit expansion, offering medicinal chemists a valuable tool for drug discovery developments. We find that SynLlama, even without training on external building blocks, can effectively generalize to unseen yet purchasable building blocks, meaning that its reconstruction capabilities extend to a broader synthesizable chemical space than the training data. We also demonstrate the use of SynLlama in a pharmaceutical context for synthesis planning of analog molecules and hit expansion leads for proposed inhibitors of target proteins.
Egocentric gesture recognition is a pivotal technology for enhancing natural human-computer interaction, yet traditional RGB-based solutions suffer from motion blur and illumination variations in dynamic scenarios. While event cameras show distinct advantages in handling high dynamic range with ultra-low power consumption, existing RGB-based architectures face inherent limitations in processing asynchronous event streams due to their synchronous frame-based nature. Moreover, from an egocentric perspective, event cameras record data that includes events generated by both head movements and hand gestures, thereby increasing the complexity of gesture recognition. To address this, we propose a novel network architecture specifically designed for event data processing, incorporating (1) a lightweight CNN with asymmetric depthwise convolutions to reduce parameters while preserving spatiotemporal features, (2) a plug-and-play state-space model as context block that decouples head movement noise from gesture dynamics, and (3) a parameter-free Bins-Temporal Shift Module (BSTM) that shifts features along bins and temporal dimensions to fuse sparse events efficiently. We further establish the EgoEvGesture dataset, the first large-scale dataset for egocentric gesture recognition using event cameras. Experimental results demonstrate that our method achieves 62.7% accuracy tested on unseen subjects with only 7M parameters, 3.1% higher than state-of-the-art approaches. Notable misclassifications in freestyle motions stem from high inter-personal variability and unseen test patterns differing from training data. Moreover, our approach achieved a remarkable accuracy of 97.0% on the DVS128 Gesture, demonstrating the effectiveness and generalization capability of our method on public datasets. The dataset and models are made available at https://github.com/3190105222/EgoEv_Gesture.
This paper focuses on formally verifying invariant properties of control programs both at the model and code levels. The physical process is described by an uncertain discrete-time state-space system, where the dependence of the state-space matrix-valued functions defining the system on the uncertainties can be rational. The proposed approaches make use of pointwise integral quadratic constraints (IQCs) to characterize the uncertainties affecting the behavior of the system. Various uncertainties can be characterized by pointwise IQCs, including static linear time-varying perturbations and sector-bounded nonlinearities. Using the IQC framework, a sound overapproximation of the uncertain system, which is expressible at the code level, is constructed. Tools such as Frama-C, ACSL, WP, and an Alt-Ergo plugin are employed to ensure the validity of the state and output invariant properties across both real and float models. The first proposed approach can be used to formally verify (local) invariant properties of the control code. This capability is demonstrated in a couple of examples involving gain-scheduled path-following controllers designed for an uncrewed aircraft system and an autonomous underwater vehicle. The second approach enables the verification of closed-loop invariant properties, i.e., invariant properties of the controlled system as a whole, in both real and float models, while preserving the integrity of the executable controller code. This is achieved by using ghost code attached to the control code for all elements related to the plant model with uncertainties, as the ghost code does not interfere with the executable code. The effectiveness of this approach is demonstrated in two examples on the control of a four-thruster hovercraft and the control of a two-mass rotational system.
Estimating the parameters of nonlinear block-oriented state-space models from input-output data typically involves solving a highly non-convex optimization problem, which is prone to poor local minima and slow convergence. This paper presents a computationally efficient initialization method for nonlinear linear fractional representation (NL-LFR) models using periodic data. By first inferring the latent signals and subsequently estimating the model parameters, the approach generates initial estimates for use in a later nonlinear optimization step. The proposed method shows robustness against poor local minima, and achieves a twofold error reduction compared to the state-of-the-art on a challenging benchmark dataset.
The traditional control theory and its application to basic and complex systems have reached an advanced level of maturity. This includes aerial, marine, and ground vehicles, as well as robotics, chemical, transportation, and electrical systems widely used in our daily lives. The emerging era of data-driven methods, Large Language Models (LLMs), and AI-based controllers does not indicate a weakness in well-established control theory. Instead, it aims to reduce dependence on models and uncertainties, address increasingly complex systems, and potentially achieve decision-making capabilities comparable to human-level performance. This revolution integrates knowledge from computer science, machine learning, biology, and classical control, producing promising algorithms that are yet to demonstrate widespread real-world applicability. Despite the maturity of control theory and the presence of various performance criteria, there is still a lack of standardised metrics for testing, evaluation, Verification and Validation ($V\&V$) of algorithms. This gap can lead to algorithms that, while optimal in certain aspects, may fall short of practical implementation, sparking debates within the literature. For a controller to succeed in real-world applications, it must satisfy three key categories of performance metrics: tracking quality, control effort (energy consumption), and robustness. This paper rather takes an applied perspective, proposing and consolidating standard performance criteria for testing and analysing control systems, intended for researchers and students. The proposed framework ensures the post-design applicability of a black-box algorithm, aligning with modern data analysis and $V\&V$ perspectives to prevent resource allocation to systems with limited impact or imprecise claims.