Yu Zhong

Yihan Xie, Sijing Li, Tianwei Lin et al.

Although electrocardiograms (ECG) play a dominant role in cardiovascular diagnosis and treatment, their intrinsic data forms and representational patterns pose significant challenges for medical multimodal large language models (Med-MLLMs) in achieving cross-modal semantic alignment. To address this gap, we propose Heartcare Suite, a unified ECG suite designed for dual signal-image modeling and understanding: (i) Heartcare-400K. A fine-grained ECG instruction dataset on top of our data pipeline engine--HeartAgent--by integrating high quality clinical ECG reports from top hospitals with open-source data. (ii) Heartcare-Bench. A systematic benchmark assessing performance of models in multi-perspective ECG understanding and cross-modal generalization, providing guidance for optimizing ECG comprehension models. (iii) HeartcareGPT. Built upon a structure-aware discrete tokenizer Beat, we propose Dual Stream Projection Alignment (DSPA) paradigm--a dual encoder projection alignment mechanism enabling joint optimizing and modeling native ECG signal-image within a shared feature space. HeartcareGPT achieves consistent improvements across diverse ECG understanding tasks, validating both the effectiveness of the unified modeling paradigm and the necessity of a high-quality data pipeline, and establishing a methodological foundation for extending Med-MLLMs towards physiological signal domains. Our project is available at https://github.com/ZJU4HealthCare/HeartcareGPT .

Tianwei Lin, Zhongwei Qiu, Jie Cao et al.

Medical vision-language models (VLMs) have rapidly advanced as general-purpose multimodal assistants, yet their deployment in 3D Computed Tomography (CT) analysis remains constrained by a persistent mismatch between optimization objectives and clinical rigor. Current Reinforcement Learning (RL) paradigms still rely on lexical proxy signals that induce ``\textit{Evaluation Hallucinations}'', where models optimize linguistic fluency rather than factual clinical correctness, leading to diagnostically critical errors. To bridge this gap, we introduce the \textbf{Clinical Abnormality Benchmarking Substrate (CABS)}, a structured system that decomposes radiology reports into verifiable clinical semantic units. Using CABS, we identify a ``\textit{Mechanistic Divergence}'' in standard RL, where surface-similarity rewards drive policy gradients to bypass medical facts. We therefore propose \textbf{Trajectory-Integral Feedback GRPO (TIF-GRPO)}, a novel framework integrating control-theoretic principles into policy optimization. By formulating clinical reasoning as a pseudo-temporal trajectory for anomaly discovery, TIF-GRPO regulates anatomy-aware rewards via an integral feedback loop that penalizes persistent omissions as cumulative state errors and suppresses hallucinations as excessive control effort. Experiments on 3D CT benchmarks demonstrate that our approach significantly enhances abnormality detection and clinical faithfulness, establishing a new paradigm for fine-grained regulation in medical VLMs. Our project is available at \href{https://github.com/ZJU4HealthCare/TIF-GRPO}{GitHub}.

Chengshuo Shen, Wei Zheng, Yonghua Ding et al.

Disruption prediction has made rapid progress in recent years, especially in machine learning (ML)-based methods. Understanding why a predictor makes a certain prediction can be as crucial as the prediction's accuracy for future tokamak disruption predictors. The purpose of most disruption predictors is accuracy or cross-machine capability. However, if a disruption prediction model can be interpreted, it can tell why certain samples are classified as disruption precursors. This allows us to tell the types of incoming disruption and gives us insight into the mechanism of disruption. This paper designs a disruption predictor called Interpretable Disruption Predictor based On Physics-guided feature extraction (IDP-PGFE) on J-TEXT. The prediction performance of the model is effectively improved by extracting physics-guided features. A high-performance model is required to ensure the validity of the interpretation results. The interpretability study of IDP-PGFE provides an understanding of J-TEXT disruption and is generally consistent with existing comprehension of disruption. IDP-PGFE has been applied to the disruption due to continuously increasing density towards density limit experiments on J-TEXT. The time evolution of the PGFE features contribution demonstrates that the application of ECRH triggers radiation-caused disruption, which lowers the density at disruption. While the application of RMP indeed raises the density limit in J-TEXT. The interpretability study guides intuition on the physical mechanisms of density limit disruption that RMPs affect not only the MHD instabilities but also the radiation profile, which delays density limit disruption.

Chengshuo Shen, Wei Zheng, Bihao Guo et al.

The high acquisition cost and the significant demand for disruptive discharges for data-driven disruption prediction models in future tokamaks pose an inherent contradiction in disruption prediction research. In this paper, we demonstrated a novel approach to predict disruption in a future tokamak using only a few discharges. The first step is to use the existing understanding of physics to extract physics-guided features from the diagnostic signals of each tokamak, called physics-guided feature extraction (PGFE). The second step is to align a few data from the future tokamak (target domain) and a large amount of data from existing tokamak (source domain) based on a domain adaptation algorithm called CORrelation ALignment (CORAL). It is the first attempt at applying domain adaptation in the task of disruption prediction. PGFE has been successfully applied in J-TEXT to predict disruption with excellent performance. PGFE can also reduce the data volume requirements due to extracting the less device-specific features, thereby establishing a solid foundation for cross-tokamak disruption prediction. We have further improved CORAL (supervised CORAL, S-CORAL) to enhance its appropriateness in feature alignment for the disruption prediction task. To simulate the existing and future tokamak case, we selected J-TEXT as the existing tokamak and EAST as the future tokamak, which has a large gap in the ranges of plasma parameters. The utilization of the S-CORAL improves the disruption prediction performance on future tokamak. Through interpretable analysis, we discovered that the learned knowledge of the disruption prediction model through this approach exhibits more similarities to the model trained on large data volumes of future tokamak.

Yuqian Yuan, Wenqiao Zhang, Juekai Lin et al.

Large Multimodal Models (LMMs) have achieved remarkable progress in general-purpose vision--language understanding, yet they remain limited in tasks requiring precise object-level grounding, fine-grained spatial reasoning, and controllable visual manipulation. In particular, existing systems often struggle to identify the correct instance, preserve object identity across interactions, and localize or modify designated regions with high precision. Object-centric vision provides a principled framework for addressing these challenges by promoting explicit representations and operations over visual entities, thereby extending multimodal systems from global scene understanding to object-level understanding, segmentation, editing, and generation. This paper presents a comprehensive review of recent advances at the convergence of LMMs and object-centric vision. We organize the literature into four major themes: object-centric visual understanding, object-centric referring segmentation, object-centric visual editing, and object-centric visual generation. We further summarize the key modeling paradigms, learning strategies, and evaluation protocols that support these capabilities. Finally, we discuss open challenges and future directions, including robust instance permanence, fine-grained spatial control, consistent multi-step interaction, unified cross-task modeling, and reliable benchmarking under distribution shift. We hope this paper provides a structured perspective on the development of scalable, precise, and trustworthy object-centric multimodal systems.

Xinkun Ai, Wei Zheng, Ming Zhang et al.

The full understanding of plasma disruption in tokamaks is currently lacking, and data-driven methods are extensively used for disruption prediction. However, most existing data-driven disruption predictors employ supervised learning techniques, which require labeled training data. The manual labeling of disruption precursors is a tedious and challenging task, as some precursors are difficult to accurately identify, limiting the potential of machine learning models. To address this issue, commonly used labeling methods assume that the precursor onset occurs at a fixed time before the disruption, which may not be consistent for different types of disruptions or even the same type of disruption, due to the different speeds at which plasma instabilities escalate. This leads to mislabeled samples and suboptimal performance of the supervised learning predictor. In this paper, we present a disruption prediction method based on anomaly detection that overcomes the drawbacks of unbalanced positive and negative data samples and inaccurately labeled disruption precursor samples. We demonstrate the effectiveness and reliability of anomaly detection predictors based on different algorithms on J-TEXT and EAST to evaluate the reliability of the precursor onset time inferred by the anomaly detection predictor. The precursor onset times inferred by these predictors reveal that the labeling methods have room for improvement as the onset times of different shots are not necessarily the same. Finally, we optimize precursor labeling using the onset times inferred by the anomaly detection predictor and test the optimized labels on supervised learning disruption predictors. The results on J-TEXT and EAST show that the models trained on the optimized labels outperform those trained on fixed onset time labels.

Zihan Cao, Xiao Wu, Liang-Jian Deng et al.

In image fusion tasks, images from different sources possess distinct characteristics. This has driven the development of numerous methods to explore better ways of fusing them while preserving their respective characteristics.Mamba, as a state space model, has emerged in the field of natural language processing. Recently, many studies have attempted to extend Mamba to vision tasks. However, due to the nature of images different from causal language sequences, the limited state capacity of Mamba weakens its ability to model image information. Additionally, the sequence modeling ability of Mamba is only capable of spatial information and cannot effectively capture the rich spectral information in images. Motivated by these challenges, we customize and improve the vision Mamba network designed for the image fusion task. Specifically, we propose the local-enhanced vision Mamba block, dubbed as LEVM. The LEVM block can improve local information perception of the network and simultaneously learn local and global spatial information. Furthermore, we propose the state sharing technique to enhance spatial details and integrate spatial and spectral information. Finally, the overall network is a multi-scale structure based on vision Mamba, called LE-Mamba. Extensive experiments show the proposed methods achieve state-of-the-art results on multispectral pansharpening and multispectral and hyperspectral image fusion datasets, and demonstrate the effectiveness of the proposed approach. Codes can be accessed at \url{https://github.com/294coder/Efficient-MIF}.

Yu Zhong, Xiao Wu, Liang-Jian Deng et al.

Pansharpening is a significant image fusion technique that merges the spatial content and spectral characteristics of remote sensing images to generate high-resolution multispectral images. Recently, denoising diffusion probabilistic models have been gradually applied to visual tasks, enhancing controllable image generation through low-rank adaptation (LoRA). In this paper, we introduce a spatial-spectral integrated diffusion model for the remote sensing pansharpening task, called SSDiff, which considers the pansharpening process as the fusion process of spatial and spectral components from the perspective of subspace decomposition. Specifically, SSDiff utilizes spatial and spectral branches to learn spatial details and spectral features separately, then employs a designed alternating projection fusion module (APFM) to accomplish the fusion. Furthermore, we propose a frequency modulation inter-branch module (FMIM) to modulate the frequency distribution between branches. The two components of SSDiff can perform favorably against the APFM when utilizing a LoRA-like branch-wise alternative fine-tuning method. It refines SSDiff to capture component-discriminating features more sufficiently. Finally, extensive experiments on four commonly used datasets, i.e., WorldView-3, WorldView-2, GaoFen-2, and QuickBird, demonstrate the superiority of SSDiff both visually and quantitatively. The code will be made open source after possible acceptance.

Zihan Cao, Yu Zhong, Ziqi Wang et al.

Image fusion, a fundamental low-level vision task, aims to integrate multiple image sequences into a single output while preserving as much information as possible from the input. However, existing methods face several significant limitations: 1) requiring task- or dataset-specific models; 2) neglecting real-world image degradations (\textit{e.g.}, noise), which causes failure when processing degraded inputs; 3) operating in pixel space, where attention mechanisms are computationally expensive; and 4) lacking user interaction capabilities. To address these challenges, we propose a unified framework for multi-task, multi-degradation, and language-guided image fusion. Our framework includes two key components: 1) a practical degradation pipeline that simulates real-world image degradations and generates interactive prompts to guide the model; 2) an all-in-one Diffusion Transformer (DiT) operating in latent space, which fuses a clean image conditioned on both the degraded inputs and the generated prompts. Furthermore, we introduce principled modifications to the original DiT architecture to better suit the fusion task. Based on this framework, we develop two versions of the model: Regression-based and Flow Matching-based variants. Extensive qualitative and quantitative experiments demonstrate that our approach effectively addresses the aforementioned limitations and outperforms previous restoration+fusion and all-in-one pipelines. Codes are available at https://github.com/294coder/MMAIF.

Zihan Cao, Yu Zhong, Liang-Jian Deng

Pansharpening, a pivotal task in remote sensing for fusing high-resolution panchromatic and multispectral imagery, has garnered significant research interest. Recent advancements employing diffusion models based on stochastic differential equations (SDEs) have demonstrated state-of-the-art performance. However, the inherent multi-step sampling process of SDEs imposes substantial computational overhead, hindering practical deployment. While existing methods adopt efficient samplers, knowledge distillation, or retraining to reduce sampling steps (e.g., from 1,000 to fewer steps), such approaches often compromise fusion quality. In this work, we propose the Optimal Transport Flow Matching (OTFM) framework, which integrates the dual formulation of unbalanced optimal transport (UOT) to achieve one-step, high-quality pansharpening. Unlike conventional OT formulations that enforce rigid distribution alignment, UOT relaxes marginal constraints to enhance modeling flexibility, accommodating the intrinsic spectral and spatial disparities in remote sensing data. Furthermore, we incorporate task-specific regularization into the UOT objective, enhancing the robustness of the flow model. The OTFM framework enables simulation-free training and single-step inference while maintaining strict adherence to pansharpening constraints. Experimental evaluations across multiple datasets demonstrate that OTFM matches or exceeds the performance of previous regression-based models and leading diffusion-based methods while only needing one sampling step. Codes are available at https://github.com/294coder/PAN-OTFM.

Jiebin Yan, Yu Zhong, Yuming Fang et al.

Semantic segmentation is an extensively studied task in computer vision, with numerous methods proposed every year. Thanks to the advent of deep learning in semantic segmentation, the performance on existing benchmarks is close to saturation. A natural question then arises: Does the superior performance on the closed (and frequently re-used) test sets transfer to the open visual world with unconstrained variations? In this paper, we take steps toward answering the question by exposing failures of existing semantic segmentation methods in the open visual world under the constraint of very limited human labeling effort. Inspired by previous research on model falsification, we start from an arbitrarily large image set, and automatically sample a small image set by MAximizing the Discrepancy (MAD) between two segmentation methods. The selected images have the greatest potential in falsifying either (or both) of the two methods. We also explicitly enforce several conditions to diversify the exposed failures, corresponding to different underlying root causes. A segmentation method, whose failures are more difficult to be exposed in the MAD competition, is considered better. We conduct a thorough MAD diagnosis of ten PASCAL VOC semantic segmentation algorithms. With detailed analysis of experimental results, we point out strengths and weaknesses of the competing algorithms, as well as potential research directions for further advancement in semantic segmentation. The codes are publicly available at \url{https://github.com/QTJiebin/MAD_Segmentation}.

Haoyu Zheng, Wenqiao Zhang, Zheqi Lv et al.

Diffusion-based text-to-image (T2I) models have demonstrated remarkable results in global video editing tasks. However, their focus is primarily on global video modifications, and achieving desired attribute-specific changes remains a challenging task, specifically in multi-attribute editing (MAE) in video. Contemporary video editing approaches either require extensive fine-tuning or rely on additional networks (such as ControlNet) for modeling multi-object appearances, yet they remain in their infancy, offering only coarse-grained MAE solutions. In this paper, we present MAKIMA, a tuning-free MAE framework built upon pretrained T2I models for open-domain video editing. Our approach preserves video structure and appearance information by incorporating attention maps and features from the inversion process during denoising. To facilitate precise editing of multiple attributes, we introduce mask-guided attention modulation, enhancing correlations between spatially corresponding tokens and suppressing cross-attribute interference in both self-attention and cross-attention layers. To balance video frame generation quality and efficiency, we implement consistent feature propagation, which generates frame sequences by editing keyframes and propagating their features throughout the sequence. Extensive experiments demonstrate that MAKIMA outperforms existing baselines in open-domain multi-attribute video editing tasks, achieving superior results in both editing accuracy and temporal consistency while maintaining computational efficiency.

Yu Zhong, Zihao Zhang, Rui Zhang et al.

Vision-and-Language Navigation (VLN) requires an agent to dynamically explore complex 3D environments following human instructions. Recent research underscores the potential of harnessing large language models (LLMs) for VLN, given their commonsense knowledge and general reasoning capabilities. Despite their strengths, a substantial gap in task completion performance persists between LLM-based approaches and domain experts, as LLMs inherently struggle to comprehend real-world spatial correlations precisely. Additionally, introducing LLMs is accompanied with substantial computational cost and inference latency. To address these issues, we propose a novel dual-process thinking framework dubbed R3, integrating LLMs' generalization capabilities with VLN-specific expertise in a zero-shot manner. The framework comprises three core modules: Runner, Ruminator, and Regulator. The Runner is a lightweight transformer-based expert model that ensures efficient and accurate navigation under regular circumstances. The Ruminator employs a powerful multimodal LLM as the backbone and adopts chain-of-thought (CoT) prompting to elicit structured reasoning. The Regulator monitors the navigation progress and controls the appropriate thinking mode according to three criteria, integrating Runner and Ruminator harmoniously. Experimental results illustrate that R3 significantly outperforms other state-of-the-art methods, exceeding 3.28% and 3.30% in SPL and RGSPL respectively on the REVERIE benchmark. This pronounced enhancement highlights the effectiveness of our method in handling challenging VLN tasks.

Dilshod Azizov, Muhammad Arslan Manzoor, Velibor Bojkovic et al.

Deep learning has fundamentally reshaped the landscape of artificial intelligence over the past decade, enabling remarkable achievements across diverse domains. At the heart of these developments lie multi-layered neural network architectures that excel at automatic feature extraction, leading to significant improvements in machine learning tasks. To demystify these advances and offer accessible guidance, we present a comprehensive overview of the most influential deep learning algorithms selected through a broad-based survey of the field. Our discussion centers on pivotal architectures, including Residual Networks, Transformers, Generative Adversarial Networks, Variational Autoencoders, Graph Neural Networks, Contrastive Language-Image Pre-training, and Diffusion models. We detail their historical context, highlight their mathematical foundations and algorithmic principles, and examine subsequent variants, extensions, and practical considerations such as training methodologies, normalization techniques, and learning rate schedules. Beyond historical and technical insights, we also address their applications, challenges, and potential research directions. This survey aims to serve as a practical manual for both newcomers seeking an entry point into cutting-edge deep learning methods and experienced researchers transitioning into this rapidly evolving domain.

Yu Zhong, Rui Zhang, Zihao Zhang et al.

Vision-and-Language Navigation (VLN) is a challenging task that requires an agent to navigate through photorealistic environments following natural-language instructions. One main obstacle existing in VLN is data scarcity, leading to poor generalization performance over unseen environments. Though data argumentation is a promising way for scaling up the dataset, how to generate VLN data both diverse and world-consistent remains problematic. To cope with this issue, we propose the world-consistent data generation (WCGEN), an efficacious data-augmentation framework satisfying both diversity and world-consistency, aimed at enhancing the generalization of agents to novel environments. Roughly, our framework consists of two stages, the trajectory stage which leverages a point-cloud based technique to ensure spatial coherency among viewpoints, and the viewpoint stage which adopts a novel angle synthesis method to guarantee spatial and wraparound consistency within the entire observation. By accurately predicting viewpoint changes with 3D knowledge, our approach maintains the world-consistency during the generation procedure. Experiments on a wide range of datasets verify the effectiveness of our method, demonstrating that our data augmentation strategy enables agents to achieve new state-of-the-art results on all navigation tasks, and is capable of enhancing the VLN agents' generalization ability to unseen environments.

Shaojie Tang, Penpen Miao, Xingyu Gao et al.

A method was proposed for the point cloud-based registration and image fusion between cardiac single photon emission computed tomography (SPECT) myocardial perfusion images (MPI) and cardiac computed tomography angiograms (CTA). Firstly, the left ventricle (LV) epicardial regions (LVERs) in SPECT and CTA images were segmented by using different U-Net neural networks trained to generate the point clouds of the LV epicardial contours (LVECs). Secondly, according to the characteristics of cardiac anatomy, the special points of anterior and posterior interventricular grooves (APIGs) were manually marked in both SPECT and CTA image volumes. Thirdly, we developed an in-house program for coarsely registering the special points of APIGs to ensure a correct cardiac orientation alignment between SPECT and CTA images. Fourthly, we employed ICP, SICP or CPD algorithm to achieve a fine registration for the point clouds (together with the special points of APIGs) of the LV epicardial surfaces (LVERs) in SPECT and CTA images. Finally, the image fusion between SPECT and CTA was realized after the fine registration. The experimental results showed that the cardiac orientation was aligned well and the mean distance error of the optimal registration method (CPD with affine transform) was consistently less than 3 mm. The proposed method could effectively fuse the structures from cardiac CTA and SPECT functional images, and demonstrated a potential in assisting in accurate diagnosis of cardiac diseases by combining complementary advantages of the two imaging modalities.

Ben Hutchinson, Vinodkumar Prabhakaran, Emily Denton et al.

Building equitable and inclusive NLP technologies demands consideration of whether and how social attitudes are represented in ML models. In particular, representations encoded in models often inadvertently perpetuate undesirable social biases from the data on which they are trained. In this paper, we present evidence of such undesirable biases towards mentions of disability in two different English language models: toxicity prediction and sentiment analysis. Next, we demonstrate that the neural embeddings that are the critical first step in most NLP pipelines similarly contain undesirable biases towards mentions of disability. We end by highlighting topical biases in the discourse about disability which may contribute to the observed model biases; for instance, gun violence, homelessness, and drug addiction are over-represented in texts discussing mental illness.

Nick Moran, Dan Schmidt, Yu Zhong et al.

We present a method for training a neural network to perform image denoising without access to clean training examples or access to paired noisy training examples. Our method requires only a single noisy realization of each training example and a statistical model of the noise distribution, and is applicable to a wide variety of noise models, including spatially structured noise. Our model produces results which are competitive with other learned methods which require richer training data, and outperforms traditional non-learned denoising methods. We present derivations of our method for arbitrary additive noise, an improvement specific to Gaussian additive noise, and an extension to multiplicative Bernoulli noise.

Ruihuan Du, Yu Zhong, Harikesh Nair et al.

This paper describes a practical system for Multi Touch Attribution (MTA) for use by a publisher of digital ads. We developed this system for JD.com, an eCommerce company, which is also a publisher of digital ads in China. The approach has two steps. The first step ('response modeling') fits a user-level model for purchase of a product as a function of the user's exposure to ads. The second ('credit allocation') uses the fitted model to allocate the incremental part of the observed purchase due to advertising, to the ads the user is exposed to over the previous T days. To implement step one, we train a Recurrent Neural Network (RNN) on user-level conversion and exposure data. The RNN has the advantage of flexibly handling the sequential dependence in the data in a semi-parametric way. The specific RNN formulation we implement captures the impact of advertising intensity, timing, competition, and user-heterogeneity, which are known to be relevant to ad-response. To implement step two, we compute Shapley Values, which have the advantage of having axiomatic foundations and satisfying fairness considerations. The specific formulation of the Shapley Value we implement respects incrementality by allocating the overall incremental improvement in conversion to the exposed ads, while handling the sequence-dependence of exposures on the observed outcomes. The system is under production at JD.com, and scales to handle the high dimensionality of the problem on the platform (attribution of the orders of about 300M users, for roughly 160K brands, across 200+ ad-types, served about 80B ad-impressions over a typical 15-day period).

Yu Zhong, Gil Ettinger

Deep learning techniques have demonstrated significant capacity in modeling some of the most challenging real world problems of high complexity. Despite the popularity of deep models, we still strive to better understand the underlying mechanism that drives their success. Motivated by observations that neurons in trained deep nets predict attributes indirectly related to the training tasks, we recognize that a deep network learns representations more general than the task at hand to disentangle impacts of multiple confounding factors governing the data, in order to isolate the effects of the concerning factors and optimize a given objective. Consequently, we propose a general framework to augment training of deep models with information on auxiliary explanatory data variables, in an effort to boost this disentanglement and train deep networks that comprehend the data interactions and distributions more accurately, and thus improve their generalizability. We incorporate information on prominent auxiliary explanatory factors of the data population into existing architectures as secondary objective/loss blocks that take inputs from hidden layers during training. Once trained, these secondary circuits can be removed to leave a model with the same architecture as the original, but more generalizable and discerning thanks to its comprehension of data interactions. Since pose is one of the most dominant confounding factors for object recognition, we apply this principle to instantiate a pose-aware deep convolutional neural network and demonstrate that auxiliary pose information indeed improves the classification accuracy in our experiments on SAR target classification tasks.

Yu Zhong

Efficient indexing and searching of high dimensional data has been an area of active research due to the growing exploitation of high dimensional data and the vulnerability of traditional search methods to the curse of dimensionality. This paper presents a new approach for fast and effective searching and indexing of high dimensional features using random partitions of the feature space. Experiments on both handwritten digits and 3-D shape descriptors have shown the proposed algorithm to be highly effective and efficient in indexing and searching real data sets of several hundred dimensions. We also compare its performance to that of the state-of-the-art locality sensitive hashing algorithm.