arXiv subjects

Tao Zhou

Publications and source records attributed to Tao Zhou.

At least 109 records · Page 6Linked to original sources

arXiv2025

Can General-Purpose Omnimodels Compete with Specialists? A Case Study in Medical Image Segmentation

The emergence of powerful, general-purpose omnimodels capable of processing diverse data modalities has raised a critical question: can these ``jack-of-all-trades'' systems perform on par with highly specialized models in knowledge-intensive domains? This work investigates this question within the high-stakes field of medical image segmentation. We conduct a comparative study analyzing the zero-shot performance of a state-of-the-art omnimodel (Gemini, the ``Nano Banana'' model) against domain-specific deep learning models on three distinct tasks: polyp (endoscopy), retinal vessel (fundus), and breast tumor segmentation (ultrasound). Our study focuses on performance at the extremes by curating subsets of the ``easiest'' and ``hardest'' cases based on the specialist models' accuracy. Our findings reveal a nuanced and task-dependent landscape. For polyp and breast tumor segmentation, specialist models excel on easy samples, but the omnimodel demonstrates greater robustness on hard samples where specialists fail catastrophically. Conversely, for the fine-grained task of retinal vessel segmentation, the specialist model maintains superior performance across both easy and hard cases. Intriguingly, qualitative analysis suggests omnimodels may possess higher sensitivity, identifying subtle anatomical features missed by human annotators. Our results indicate that while current omnimodels are not yet a universal replacement for specialists, their unique strengths suggest a potential complementary role with specialist models, particularly in enhancing robustness on challenging edge cases.

eess.IV↗

arXiv2025

Energy based diffusion generator for efficient sampling of Boltzmann distributions

Sampling from Boltzmann distributions, particularly those tied to high dimensional and complex energy functions, poses a significant challenge in many fields. In this work, we present the Energy-Based Diffusion Generator (EDG), a novel approach that integrates ideas from variational autoencoders and diffusion models. EDG uses a decoder to generate Boltzmann-distributed samples from simple latent variables, and a diffusion-based encoder to estimate the Kullback-Leibler divergence to the target distribution. Notably, EDG is simulation-free, eliminating the need to solve ordinary or stochastic differential equations during training. Furthermore, by removing constraints such as bijectivity in the decoder, EDG allows for flexible network design. Through empirical evaluation, we demonstrate the superior performance of EDG across a variety of sampling tasks with complex target distributions, outperforming existing methods.

cs.LG↗

arXiv2025

Cross-Domain Attribute Alignment with CLIP: A Rehearsal-Free Approach for Class-Incremental Unsupervised Domain Adaptation

Class-Incremental Unsupervised Domain Adaptation (CI-UDA) aims to adapt a model from a labeled source domain to an unlabeled target domain, where the sets of potential target classes appearing at different time steps are disjoint and are subsets of the source classes. The key to solving this problem lies in avoiding catastrophic forgetting of knowledge about previous target classes during continuously mitigating the domain shift. Most previous works cumbersomely combine two technical components. On one hand, they need to store and utilize rehearsal target sample from previous time steps to avoid catastrophic forgetting; on the other hand, they perform alignment only between classes shared across domains at each time step. Consequently, the memory will continuously increase and the asymmetric alignment may inevitably result in knowledge forgetting. In this paper, we propose to mine and preserve domain-invariant and class-agnostic knowledge to facilitate the CI-UDA task. Specifically, via using CLIP, we extract the class-agnostic properties which we name as "attribute". In our framework, we learn a "key-value" pair to represent an attribute, where the key corresponds to the visual prototype and the value is the textual prompt. We maintain two attribute dictionaries, each corresponding to a different domain. Then we perform attribute alignment across domains to mitigate the domain shift, via encouraging visual attention consistency and prediction consistency. Through attribute modeling and cross-domain alignment, we effectively reduce catastrophic knowledge forgetting while mitigating the domain shift, in a rehearsal-free way. Experiments on three CI-UDA benchmarks demonstrate that our method outperforms previous state-of-the-art methods and effectively alleviates catastrophic forgetting. Code is available at https://github.com/RyunMi/VisTA.

cs.CV↗

arXiv2025

Interpretable Data Mining of Follicular Thyroid Cancer Ultrasound Features Using Enhanced Association Rules

Purpose: Thyroid cancer has been a common cancer. Papillary thyroid cancer and follicular thyroid cancer are the two most common types of thyroid cancer. Follicular thyroid cancer lacks distinctive ultrasound signs and is more difficult to diagnose preoperatively than the more prevalent papillary thyroid cancer, and the clinical studies associated with it are less well established. We aimed to analyze the clinical data of follicular thyroid cancer based on a novel data mining tool to identify some clinical indications that may help in preoperative diagnosis. Methods: We performed a retrospective analysis based on case data collected by the Department of General Surgery of Peking University Third Hospital between 2010 and 2023. Unlike traditional statistical methods, we improved the association rule mining, a classical data mining method, and proposed new analytical metrics reflecting the malignant association between clinical indications and cancer with the help of the idea of SHAP method in interpretable machine learning. Results: The dataset was preprocessed to contain 1673 cases (in terms of nodes rather than patients), of which 1414 were benign and 259 were malignant nodes. Our analysis pointed out that in addition to some common indicators (e.g., irregular or lobulated nodal margins, uneven thickness halo, hypoechogenicity), there were also some indicators with strong malignant associations, such as nodule-in-nodule pattern, trabecular pattern, and low TSH scores. In addition, our results suggest that the combination of Hashimoto's thyroiditis may also have a strong malignant association. Conclusion: In the preoperative diagnosis of nodules suspected of follicular thyroid cancer, multiple clinical indications should be considered for a more accurate diagnosis. The diverse malignant associations identified in our study may serve as a reference for clinicians in related fields.

cs.LG↗

arXiv2025

Operating advanced scientific instruments with AI agents that learn on the job

Advanced scientific user facilities, such as next generation X-ray light sources and self-driving laboratories, are revolutionizing scientific discovery by automating routine tasks and enabling rapid experimentation and characterizations. However, these facilities must continuously evolve to support new experimental workflows, adapt to diverse user projects, and meet growing demands for more intricate instruments and experiments. This continuous development introduces significant operational complexity, necessitating a focus on usability, reproducibility, and intuitive human-instrument interaction. In this work, we explore the integration of agentic AI, powered by Large Language Models (LLMs), as a transformative tool to achieve this goal. We present our approach to developing a human-in-the-loop pipeline for operating advanced instruments including an X-ray nanoprobe beamline and an autonomous robotic station dedicated to the design and characterization of materials. Specifically, we evaluate the potential of various LLMs as trainable scientific assistants for orchestrating complex, multi-task workflows, which also include multimodal data, optimizing their performance through optional human input and iterative learning. We demonstrate the ability of AI agents to bridge the gap between advanced automation and user-friendly operation, paving the way for more adaptable and intelligent scientific facilities.

physics.ins-det↗

arXiv2025

Hierarchical Spatio-temporal Segmentation Network for Ejection Fraction Estimation in Echocardiography Videos

Automated segmentation of the left ventricular endocardium in echocardiography videos is a key research area in cardiology. It aims to provide accurate assessment of cardiac structure and function through Ejection Fraction (EF) estimation. Although existing studies have achieved good segmentation performance, their results do not perform well in EF estimation. In this paper, we propose a Hierarchical Spatio-temporal Segmentation Network (\ourmodel) for echocardiography video, aiming to improve EF estimation accuracy by synergizing local detail modeling with global dynamic perception. The network employs a hierarchical design, with low-level stages using convolutional networks to process single-frame images and preserve details, while high-level stages utilize the Mamba architecture to capture spatio-temporal relationships. The hierarchical design balances single-frame and multi-frame processing, avoiding issues such as local error accumulation when relying solely on single frames or neglecting details when using only multi-frame data. To overcome local spatio-temporal limitations, we propose the Spatio-temporal Cross Scan (STCS) module, which integrates long-range context through skip scanning across frames and positions. This approach helps mitigate EF calculation biases caused by ultrasound image noise and other factors.

cs.CV↗

arXiv2025

Deconstructing Mobility Segregation: A Network Analysis of Racialized Flows in Pandemic-Era NYC

Urban segregation research has long relied on residential patterns, yet growing evidence suggests that racial/ethnic segregation also manifests systematically in mobility behaviors. Leveraging anonymized mobile device data from New York City before and during the COVID-19 pandemic, we develop a network-analytic framework to dissect mobility segregation in racialized flow networks. We examine citywide racial mixing patterns through mixing matrices and assortativity indices, revealing persistent diagonal dominance where intra-group flows constituted 69.27% of total movements. Crucially, we develop a novel dual-metric framework that reconceptualizes mobility segregation as two interlocking dimensions: structural segregation-passive exposure patterns driven by residential clustering, and preferential segregation-active homophily in mobility choices beyond spatial constraints. Our gravity-adjusted indices reveal that racial divisions transcend residential clustering-all racial groups exhibit active homophily after spatial adjustments, with particularly severe isolation during the pandemic. Findings highlight how pandemic restrictions disproportionately amplified asymmetric isolation, with minority communities experiencing steeper declines in access to White-dominated spaces. Finally, we propose a Homophily Gravity Model with racial similarity parameter, which significantly improves the prediction accuracy of race-specific mobility flows and more accurately reproduces intra-group mobility preferences and directional exposure patterns.Overall, this study redefines mobility segregation as a multidimensional phenomenon, where structural constraints, active preferences, and crisis responses compound to reshape urban racial inequality.

physics.soc-ph↗

arXiv2025

Deep learning for the semi-classical limit of the Schrödinger equation

In this paper, we integrate neural networks and Gaussian wave packets to numerically solve the Schrödinger equation with a smooth potential near the semi-classical limit. Our focus is not only on accurately obtaining solutions when the non-dimensional Planck's constant, $\varepsilon$, is small, but also on constructing an operator that maps initial values to solutions for the Schrödinger equation with multiscale properties. Using Gaussian wave packets framework, we first reformulate the Schrödinger equation as a system of ordinary differential equations. For a single initial condition, we solve the resulting system using PINNs or MscaleDNNs. Numerical simulations indicate that MscaleDNNs outperform PINNs, improving accuracy by one to two orders of magnitude. When dealing with a set of initial conditions, we adopt an operator-learning approach, such as physics-informed DeepONets. Numerical examples validate the effectiveness of physics-informed DeepONets with Gaussian wave packets in accurately mapping initial conditions to solutions.

physics.comp-ph↗

arXiv2025

Serial Over Parallel: Learning Continual Unification for Multi-Modal Visual Object Tracking and Benchmarking

Unifying multiple multi-modal visual object tracking (MMVOT) tasks draws increasing attention due to the complementary nature of different modalities in building robust tracking systems. Existing practices mix all data sensor types in a single training procedure, structuring a parallel paradigm from the data-centric perspective and aiming for a global optimum on the joint distribution of the involved tasks. However, the absence of a unified benchmark where all types of data coexist forces evaluations on separated benchmarks, causing \textit{inconsistency} between training and testing, thus leading to performance \textit{degradation}. To address these issues, this work advances in two aspects: \ding{182} A unified benchmark, coined as UniBench300, is introduced to bridge the inconsistency by incorporating multiple task data, reducing inference passes from three to one and cutting time consumption by 27\%. \ding{183} The unification process is reformulated in a serial format, progressively integrating new tasks. In this way, the performance degradation can be specified as knowledge forgetting of previous tasks, which naturally aligns with the philosophy of continual learning (CL), motivating further exploration of injecting CL into the unification process. Extensive experiments conducted on two baselines and four benchmarks demonstrate the significance of UniBench300 and the superiority of CL in supporting a stable unification process. Moreover, while conducting dedicated analyses, the performance degradation is found to be negatively correlated with network capacity. Additionally, modality discrepancies contribute to varying degradation levels across tasks (RGBT > RGBD > RGBE in MMVOT), offering valuable insights for future multi-modal vision research. Source codes and the proposed benchmark is available at \textit{https://github.com/Zhangyong-Tang/UniBench300}.

cs.CV↗

arXiv2025

Performances and Correlations of Centrality Measures in Complex Networks

Numerous centrality measures have been proposed to evaluate the importance of nodes in networks, yet comparative analyses of these measures remain limited. Based on 80 real-world networks, we conducted an empirical analysis of 16 representative centrality measures. In general, there exists a moderate to high level of correlation between node rankings derived from different measures. We identified two distinct communities: one comprising 4 measures and the other 7 measures. Measures within the same community exhibit exceptionally strong pairwise correlations. In contrast, the remaining five measures display markedly different behaviors, showing weak correlations not only among themselves but also with the other measures. This suggests that each of these five measures likely captures unique properties of node importance. Further analysis reveals that the distribution patterns of the most influential nodes identified by different centrality measures vary significantly: some measures tend to cluster influential nodes closely together, while others disperse them across distant locations within the network. Using the epidemic spreading model, we found that LocalRank, Subgraph Centrality, and Katz Centrality perform best in identifying the most influential single node, whereas Leverage Centrality, Collective Influence, and Cycle Ratio excel in identifying the most influential node sets. Overall, measures that identify influential nodes with larger topological distances between them tend to perform better in detecting influential node sets. Interestingly, despite being applied to the same dynamical process, when using two seemingly similar tasks, identifying influential nodes versus identifying influential node sets, to rank the performances of the 16 centrality measures, the resulting rankings are negatively correlated.

stat.OT↗

arXiv2025

Uncertainty-aware Cross-training for Semi-supervised Medical Image Segmentation

Semi-supervised learning has gained considerable popularity in medical image segmentation tasks due to its capability to reduce reliance on expert-examined annotations. Several mean-teacher (MT) based semi-supervised methods utilize consistency regularization to effectively leverage valuable information from unlabeled data. However, these methods often heavily rely on the student model and overlook the potential impact of cognitive biases within the model. Furthermore, some methods employ co-training using pseudo-labels derived from different inputs, yet generating high-confidence pseudo-labels from perturbed inputs during training remains a significant challenge. In this paper, we propose an Uncertainty-aware Cross-training framework for semi-supervised medical image Segmentation (UC-Seg). Our UC-Seg framework incorporates two distinct subnets to effectively explore and leverage the correlation between them, thereby mitigating cognitive biases within the model. Specifically, we present a Cross-subnet Consistency Preservation (CCP) strategy to enhance feature representation capability and ensure feature consistency across the two subnets. This strategy enables each subnet to correct its own biases and learn shared semantics from both labeled and unlabeled data. Additionally, we propose an Uncertainty-aware Pseudo-label Generation (UPG) component that leverages segmentation results and corresponding uncertainty maps from both subnets to generate high-confidence pseudo-labels. We extensively evaluate the proposed UC-Seg on various medical image segmentation tasks involving different modality images, such as MRI, CT, ultrasound, colonoscopy, and so on. The results demonstrate that our method achieves superior segmentation accuracy and generalization performance compared to other state-of-the-art semi-supervised methods. Our code will be released at https://github.com/taozh2017/UCSeg.

cs.CV↗

arXiv2025

An RBC-MsUQ Framework for Red Blood Cell Morpho-Mechanics

Characterizing the morpho-mechanical properties of red blood cells (RBCs) is crucial for understanding microvascular transport mechanisms and cellular pathophysiological processes, yet current computational models are constrained by multi-source uncertainties including cross-platform experimental discrepancies and parameter identification stochasticity. We present RBC-MsUQ, a novel multi-stage uncertainty quantification framework tailored for RBCs. It integrates hierarchical Bayesian inference with diverse experimental datasets, establishing prior distributions for RBC parameters via microscopic simulations and literature-derived data. A dynamic annealing technique defines stress-free baselines, while deep neural network surrogates, optimized through sensitivity analysis, achieve sub-10$^{-2}$ prediction errors for efficient simulation approximation. Its two-stage hierarchical inference architecture constrains geometric and shear modulus parameters using stress-free state and stretching data in Stage I and enables full-parameter identification via membrane fluctuation and relaxation tests in Stage II. Applied to healthy and malaria-infected RBCs, the RBC-MsUQ framework produces statistically robust posterior distributions, revealing increased stiffness and viscosity in pathological cells. Quantitative model-experiment validation demonstrates that RBC-MsUQ effectively mitigates uncertainties through cross-platform data fusion, overcoming the critical limitations of existing computational approaches. The RBC-MsUQ framework thus provides a systematic paradigm for studying RBC properties and advancing cellular mechanics and biomedical engineering.

physics.bio-ph↗

arXiv2025

Finite time blowup for Keller-Segel equation with logistic damping in three dimensions

The Keller-Segel equation, a classical chemotaxis model, and many of its variants have been extensively studied for decades. In this work, we focus on 3D Keller-Segel equation with a quadratic logistic damping term $-μρ^2$ (modeling density-dependent mortality rate) and show the existence of finite-time blowup solutions with nonnegative density and finite mass for any $μ\in \big[0,\frac{1}{3}\big)$. This range of $μ$ is sharp; for $μ\ge \frac{1}{3}$, the logistic damping effect suppresses the blowup as shown in [Kang-Stevens, 2016] and [Tello-Winkler, 2007]. A key ingredient is to construct a self-similar blowup solution to a related aggregation equation as an approximate solution, with subcritical scaling relative to the original model. Based on this construction, we employ a robust weighted $L^2$ method to prove the stability of this approximate solution, where modulation ODEs are introduced to enforce local vanishing conditions for the perturbation lying in a singular-weighted $L^2$ space. As a byproduct, we exhibit a new family of type I blowup mechanisms for the classical 3D Keller-Segel equation.

math.AP↗

arXiv2025

LVM-GP: Uncertainty-Aware PDE Solver via coupling latent variable model and Gaussian process

We propose a novel probabilistic framework, termed LVM-GP, for uncertainty quantification in solving forward and inverse partial differential equations (PDEs) with noisy data. The core idea is to construct a stochastic mapping from the input to a high-dimensional latent representation, enabling uncertainty-aware prediction of the solution. Specifically, the architecture consists of a confidence-aware encoder and a probabilistic decoder. The encoder implements a high-dimensional latent variable model based on a Gaussian process (LVM-GP), where the latent representation is constructed by interpolating between a learnable deterministic feature and a Gaussian process prior, with the interpolation strength adaptively controlled by a confidence function learned from data. The decoder defines a conditional Gaussian distribution over the solution field, where the mean is predicted by a neural operator applied to the latent representation, allowing the model to learn flexible function-to-function mapping. Moreover, physical laws are enforced as soft constraints in the loss function to ensure consistency with the underlying PDE structure. Compared to existing approaches such as Bayesian physics-informed neural networks (B-PINNs) and deep ensembles, the proposed framework can efficiently capture functional dependencies via merging a latent Gaussian process and neural operator, resulting in competitive predictive accuracy and robust uncertainty quantification. Numerical experiments demonstrate the effectiveness and reliability of the method.

stat.ML↗

arXiv2025

Effective perpendicular electric field as a probe for interlayer pairing in ambient-pressure superconducting La$_{2.85}$Pr$_{0.15}$Ni$_{2}$O$_{7}$ thin films

Recent angle-resolved photoemission spectroscopy (ARPES) experiments on La$_{2.85}$Pr$_{0.15}$Ni$_2$O$_7$ thin films have revealed a superconducting gap near the diagonal direction for the $d_{x^2-y^2}$ orbital, confined to a narrow momentum region. Our numerical calculations indicate that interlayer pairing is dominant, resulting in an $s_\pm$-wave symmetry. We propose employing perpendicular electric fields, a practical method in thin film systems, to verify the pairing symmetry. Our calculations predict that such fields will induce Fermi arcs or nodal points and significantly modulate the gap near the diagonal, with the modulated region overlapping the area probed by ARPES. This approach provides an effective tool for probing pairing symmetry in bilayer nickelate thin films, especially given current limitations in experimental techniques.

cond-mat.supr-con↗

arXiv2025

Bounded KRnet and its applications to density estimation and approximation

In this paper, we develop an invertible mapping, called B-KRnet, on a bounded domain and apply it to density estimation/approximation for data or the solutions of PDEs such as the Fokker-Planck equation and the Keller-Segel equation. Similar to KRnet, B-KRnet consists of a series of coupling layers with progressively fewer active transformation dimensions, inspired by the triangular structure of the Knothe-Rosenblatt (KR) rearrangement. The main difference between B-KRnet and KRnet is that B-KRnet is defined on a hypercube while KRnet is defined on the whole space, in other words, a new mechanism is introduced in B-KRnet to maintain the exact invertibility. Using B-KRnet as a transport map, we obtain an explicit probability density function (PDF) model that corresponds to the pushforward of a base (uniform) distribution on the hypercube. It can be directly applied to density estimation when only data are available. By coupling KRnet and B-KRnet, we define a deep generative model on a high-dimensional domain where some dimensions are bounded and other dimensions are unbounded. A typical case is the solution of the stationary kinetic Fokker-Planck equation, which is a PDF of position and momentum. Based on B-KRnet, we develop an adaptive learning approach to approximate partial differential equations whose solutions are PDFs or can be treated as PDFs. A variety of numerical experiments is presented to demonstrate the effectiveness of B-KRnet.

cs.LG↗

arXiv2025

Superlinear gradient growth for 2D Euler equation without boundary

We consider the vorticity gradient growth of solutions to the two-dimensional Euler equations in domains without boundary, namely in the torus $\mathbb{T}^{2}$ and the whole plane $\mathbb{R}^{2}$. In the torus, whenever we have a steady state $ω^*$ that is orbitally stable up to a translation and has a saddle point, we construct ${\tildeω}_0 \in C^\infty(\mathbb{T}^2)$ that is arbitrarily close to $ω^*$ in $L^2$, such that superlinear growth of the vorticity gradient occurs for an open set of smooth initial data around ${\tildeω}_0$. This seems to be the first superlinear growth result which holds for an open set of smooth initial data (and does not require any symmetry assumptions on the initial vorticity). Furthermore, we obtain the first superlinear growth result for smooth and compactly supported vorticity in the plane, using perturbations of the Lamb-Chaplygin dipole.

math.AP↗

arXiv2025

DONUT: Physics-aware Machine Learning for Real-time X-ray Nanodiffraction Analysis

Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample's local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.

cs.LG↗