Xiaoxiao Sun

Junzhe Huang, Xiaoxiao Sun, Yan Yang et al.

Using multimodal foundation models to analyze table images is a high-value yet challenging application in consumer and enterprise scenarios. Despite its importance, current evaluations rely largely on structured-text tables or clean rendered images, leaving the visual complexity of in-the-wild table images underexplored. Such images feature varied layouts and diverse domains that demand sophisticated structural perception and numerical reasoning. To bridge this gap, we introduce WildTableBench, the first question-answering benchmark for naturally occurring table images from real-world settings. WildTableBench comprises 402 high-information-density table images collected from online forums and websites across diverse domains, together with 928 manually annotated and verified questions spanning 17 subtypes across five categories. We evaluate 21 frontier proprietary and open-source multimodal foundation models on this benchmark. Only one model exceeds 50% accuracy, while all remaining models range from 4.1% to 49.9%. We further conduct diagnostic analyses to characterize model failures and reveal persistent weaknesses in structural perception and reasoning. These results and analyses provide useful insights into current model capabilities and establish WildTableBench as a valuable diagnostic benchmark for table image understanding.

Jincheng Yu, Haoyang Li, Yiwen Liu et al.

Purpose: To develop an interpretable and trustworthy AI framework that combines deep learning based MRI Osteoarthritis Knee Score (MOAKS) prediction with interpretable statistical modeling to study structure-pain relationships at scale using data from the Osteoarthritis Initiative (OAI). Materials and Methods: We first developed a deep learning framework to predict MOAKS features directly from knee MRIs and incorporated conformal prediction to provide prediction uncertainty quantification. This uncertainty-aware strategy enables explicit filtering of model outputs, retaining only high-confidence MOAKS predictions at the knee level. Second, we applied a longitudinal latent class mixed model (LCMM) to examine associations between key structural abnormalities and four complementary knee pain measurements. Results: Among the three MRI-defined abnormalities (i.e., bone marrow lesions (BML), cartilage loss (CART), and meniscal extrusion (ME)), our framework substantially improved the Matthews correlation coefficient (MCC) and some other metrics. For example, MCC increased from 0.69 to 0.91 for BML, from 0.45 to 0.80 for CART, and from 0.59 to 0.89 for ME. Using these high-confidence predictions, we expanded the sample size to 2,175 knees for the LCMM analysis. Two distinct pain trajectories were identified (rapid and stable pain progression). The estimated odds ratios (95% CI) for the rapid progression group were 1.62 (1.12-2.35) for BML, 1.83 (1.24-2.70) for CART loss, and 2.50 (1.75-3.57) for ME. Conclusion: These results highlight the importance of these structural abnormalities as risk factors for pain and functional progression in osteoarthritis.

Zhongying Deng, Cheng Tang, Ziyan Huang et al. · pku

Foundation models have demonstrated remarkable success across diverse domains and tasks, primarily due to the thrive of large-scale, diverse, and high-quality datasets. However, in the field of medical imaging, the curation and assembling of such medical datasets are highly challenging due to the reliance on clinical expertise and strict ethical and privacy constraints, resulting in a scarcity of large-scale unified medical datasets and hindering the development of powerful medical foundation models. In this work, we present the largest survey to date of medical image datasets, covering over 1,000 open-access datasets with a systematic catalog of their modalities, tasks, anatomies, annotations, limitations, and potential for integration. Our analysis exposes a landscape that is modest in scale, fragmented across narrowly scoped tasks, and unevenly distributed across organs and modalities, which in turn limits the utility of existing medical image datasets for developing versatile and robust medical foundation models. To turn fragmentation into scale, we propose a metadata-driven fusion paradigm (MDFP) that integrates public datasets with shared modalities or tasks, thereby transforming multiple small data silos into larger, more coherent resources. Building on MDFP, we release an interactive discovery portal that enables end-to-end, automated medical image dataset integration, and compile all surveyed datasets into a unified, structured table that clearly summarizes their key characteristics and provides reference links, offering the community an accessible and comprehensive repository. By charting the current terrain and offering a principled path to dataset consolidation, our survey provides a practical roadmap for scaling medical imaging corpora, supporting faster data discovery, more principled dataset creation, and more capable medical foundation models.

Wenjin Hou, Wei Liu, Han Hu et al.

Multimodal Large Language Models (MLLMs) have shown remarkable proficiency on general-purpose vision-language benchmarks, reaching or even exceeding human-level performance. However, these evaluations typically rely on standard in-distribution data, leaving the robustness of MLLMs largely unexamined when faced with scenarios that defy common-sense priors. To address this gap, we introduce VIA-Bench, a challenging benchmark designed to probe model performance on visual illusions and anomalies. It includes six core categories: color illusions, motion illusions, gestalt illusions, geometric and spatial illusions, general visual illusions, and visual anomalies. Through careful human-in-the-loop review, we construct over 1K high-quality question-answer pairs that require nuanced visual reasoning. Extensive evaluation of over 20 state-of-the-art MLLMs, including proprietary, open-source, and reasoning-enhanced models, uncovers significant vulnerabilities. Notably, we find that Chain-of-Thought (CoT) reasoning offers negligible robustness, often yielding ``brittle mirages'' where the model's logic collapses under illusory stimuli. Our findings reveal a fundamental divergence between machine and human perception, suggesting that resolving such perceptual bottlenecks is critical for the advancement of artificial general intelligence. The benchmark data and code will be released.

Sharath Koorathota, Nikolas Papadopoulos, Jia Li Ma et al.

Vision Transformers (ViT) have advanced computer vision, yet their efficacy in complex tasks like driving remains less explored. This study enhances ViT by integrating human eye gaze, captured via eye-tracking, to increase prediction accuracy in driving scenarios under uncertainty in both real-world and virtual reality scenarios. First, we establish the significance of human eye gaze in left-right driving decisions, as observed in both human subjects and a ViT model. By comparing the similarity between human fixation maps and ViT attention weights, we reveal the dynamics of overlap across individual heads and layers. This overlap demonstrates that fixation data can guide the model in distributing its attention weights more effectively. We introduce the fixation-attention intersection (FAX) loss, a novel loss function that significantly improves ViT performance under high uncertainty conditions. Our results show that ViT, when trained with FAX loss, aligns its attention with human gaze patterns. This gaze-informed approach has significant potential for driver behavior analysis, as well as broader applications in human-centered AI systems, extending ViT's use to complex visual environments.

Xiaoxiao Sun, Nidham Gazagnadou, Vivek Sharma et al.

Hand-crafted image quality metrics, such as PSNR and SSIM, are commonly used to evaluate model privacy risk under reconstruction attacks. Under these metrics, reconstructed images that are determined to resemble the original one generally indicate more privacy leakage. Images determined as overall dissimilar, on the other hand, indicate higher robustness against attack. However, there is no guarantee that these metrics well reflect human opinions, which, as a judgement for model privacy leakage, are more trustworthy. In this paper, we comprehensively study the faithfulness of these hand-crafted metrics to human perception of privacy information from the reconstructed images. On 5 datasets ranging from natural images, faces, to fine-grained classes, we use 4 existing attack methods to reconstruct images from many different classification models and, for each reconstructed image, we ask multiple human annotators to assess whether this image is recognizable. Our studies reveal that the hand-crafted metrics only have a weak correlation with the human evaluation of privacy leakage and that even these metrics themselves often contradict each other. These observations suggest risks of current metrics in the community. To address this potential risk, we propose a learning-based measure called SemSim to evaluate the Semantic Similarity between the original and reconstructed images. SemSim is trained with a standard triplet loss, using an original image as an anchor, one of its recognizable reconstructed images as a positive sample, and an unrecognizable one as a negative. By training on human annotations, SemSim exhibits a greater reflection of privacy leakage on the semantic level. We show that SemSim has a significantly higher correlation with human judgment compared with existing metrics. Moreover, this strong correlation generalizes to unseen datasets, models and attack methods.

Xiaoxiao Sun, Paul Sajda

There is a growing interest in characterizing circular data found in biological systems. Such data are wide ranging and varied, from signal phase in neural recordings to nucleotide sequences in round genomes. Traditional clustering algorithms are often inadequate due to their limited ability to distinguish differences in the periodic component. Current clustering schemes that work in a polar coordinate system have limitations, such as being only angle-focused or lacking generality. To overcome these limitations, we propose a new analysis framework that utilizes projections onto a cylindrical coordinate system to better represent objects in a polar coordinate system. Using the mathematical properties of circular data, we show our approach always finds the correct clustering result within the reconstructed dataset, given sufficient periodic repetitions of the data. Our approach is generally applicable and adaptable and can be incorporated into most state-of-the-art clustering algorithms. We demonstrate on synthetic and real data that our method generates more appropriate and consistent clustering results compared to standard methods. In summary, our proposed analysis framework overcomes the limitations of existing polar coordinate-based clustering methods and provides a more accurate and efficient way to cluster circular data.

Xiaoxiao Sun, Mingyang Li, Kun yuan et al.

Large Vision-Language Models (VLMs) often answer classic visual illusions "correctly" on original images, yet persist with the same responses when illusion factors are inverted, even though the visual change is obvious to humans. This raises a fundamental question: do VLMs perceive visual changes or merely recall memorized patterns? While several studies have noted this phenomenon, the underlying causes remain unclear. To move from observations to systematic understanding, this paper introduces VI-Probe, a controllable visual-illusion framework with graded perturbations and matched visual controls (without illusion inducer) that disentangles visually grounded perception from language-driven recall. Unlike prior work that focuses on averaged accuracy, we measure stability and sensitivity using Polarity-Flip Consistency, Template Fixation Index, and an illusion multiplier normalized against matched controls. Experiments across different families reveal that response persistence arises from heterogeneous causes rather than a single mechanism. For instance, GPT-5 exhibits memory override, Claude-Opus-4.1 shows perception-memory competition, while Qwen variants suggest visual-processing limits. Our findings challenge single-cause views and motivate probing-based evaluation that measures both knowledge and sensitivity to controlled visual change. Data and code are available at https://sites.google.com/view/vi-probe/.

Xiaoxiao Sun, Xingjian Leng, Zijian Wang et al.

Analyzing model performance in various unseen environments is a critical research problem in the machine learning community. To study this problem, it is important to construct a testbed with out-of-distribution test sets that have broad coverage of environmental discrepancies. However, existing testbeds typically either have a small number of domains or are synthesized by image corruptions, hindering algorithm design that demonstrates real-world effectiveness. In this paper, we introduce CIFAR-10-Warehouse, consisting of 180 datasets collected by prompting image search engines and diffusion models in various ways. Generally sized between 300 and 8,000 images, the datasets contain natural images, cartoons, certain colors, or objects that do not naturally appear. With CIFAR-10-W, we aim to enhance the evaluation and deepen the understanding of two generalization tasks: domain generalization and model accuracy prediction in various out-of-distribution environments. We conduct extensive benchmarking and comparison experiments and show that CIFAR-10-W offers new and interesting insights inherent to these tasks. We also discuss other fields that would benefit from CIFAR-10-W.

Kun Yuan, Min Woo Sun, Zhen Chen et al.

There is a growing interest in developing strong biomedical vision-language models. A popular approach to achieve robust representations is to use web-scale scientific data. However, current biomedical vision-language pretraining typically compresses rich scientific figures and text into coarse figure-level pairs, discarding the fine-grained correspondences that clinicians actually rely on when zooming into local structures. To tackle this issue, we introduce Panel2Patch, a novel data pipeline that mines hierarchical structure from existing biomedical scientific literature, i.e., multi-panel, marker-heavy figures and their surrounding text, and converts them into multi-granular supervision. Given scientific figures and captions, Panel2Patch parses layouts, panels, and visual markers, then constructs hierarchical aligned vision-language pairs at the figure, panel, and patch levels, preserving local semantics instead of treating each figure as a single data sample. Built on this hierarchical corpus, we develop a granularity-aware pretraining strategy that unifies heterogeneous objectives from coarse didactic descriptions to fine region-focused phrases. By applying Panel2Patch to only a small set of the literature figures, we extract far more effective supervision than prior pipelines, enabling substantially better performance with less pretraining data.

Xiaoxiao Sun, Yue Yao, Shengjin Wang et al.

For object re-identification (re-ID), learning from synthetic data has become a promising strategy to cheaply acquire large-scale annotated datasets and effective models, with few privacy concerns. Many interesting research problems arise from this strategy, e.g., how to reduce the domain gap between synthetic source and real-world target. To facilitate developing more new approaches in learning from synthetic data, we introduce the Alice benchmarks, large-scale datasets providing benchmarks as well as evaluation protocols to the research community. Within the Alice benchmarks, two object re-ID tasks are offered: person and vehicle re-ID. We collected and annotated two challenging real-world target datasets: AlicePerson and AliceVehicle, captured under various illuminations, image resolutions, etc. As an important feature of our real target, the clusterability of its training set is not manually guaranteed to make it closer to a real domain adaptation test scenario. Correspondingly, we reuse existing PersonX and VehicleX as synthetic source domains. The primary goal is to train models from synthetic data that can work effectively in the real world. In this paper, we detail the settings of Alice benchmarks, provide an analysis of existing commonly-used domain adaptation methods, and discuss some interesting future directions. An online server has been set up for the community to evaluate methods conveniently and fairly. Datasets and the online server details are available at https://sites.google.com/view/alice-benchmarks.

Shengzhuang Chen, Yikai Liao, Xiaoxiao Sun et al.

The advent of the foundation model era has sparked significant research interest in leveraging pre-trained representations for continual learning (CL), yielding a series of top-performing CL methods on standard evaluation benchmarks. Nonetheless, there are growing concerns regarding potential data contamination during the pre-training stage. Furthermore, standard evaluation benchmarks, which are typically static, fail to capture the complexities of real-world CL scenarios, resulting in saturated performance. To address these issues, we describe CL on dynamic benchmarks (CLDyB), a general computational framework based on Markov decision processes for evaluating CL methods reliably. CLDyB dynamically identifies inherently difficult and algorithm-dependent tasks for the given CL methods, and determines challenging task orders using Monte Carlo tree search. Leveraging CLDyB, we first conduct a joint evaluation of multiple state-of-the-art CL methods, leading to a set of commonly challenging and generalizable task sequences where existing CL methods tend to perform poorly. We then conduct separate evaluations of individual CL methods using CLDyB, discovering their respective strengths and weaknesses. The source code and generated task sequences are publicly accessible at https://github.com/szc12153/CLDyB.

Xiaoxiao Sun, Yunzhong Hou, Hongdong Li et al.

Label-free model evaluation, or AutoEval, estimates model accuracy on unlabeled test sets, and is critical for understanding model behaviors in various unseen environments. In the absence of image labels, based on dataset representations, we estimate model performance for AutoEval with regression. On the one hand, image feature is a straightforward choice for such representations, but it hampers regression learning due to being unstructured (\ie no specific meanings for component at certain location) and of large-scale. On the other hand, previous methods adopt simple structured representations (like average confidence or average feature), but insufficient to capture the data characteristics given their limited dimensions. In this work, we take the best of both worlds and propose a new semi-structured dataset representation that is manageable for regression learning while containing rich information for AutoEval. Based on image features, we integrate distribution shapes, clusters, and representative samples for a semi-structured dataset representation. Besides the structured overall description with distribution shapes, the unstructured description with clusters and representative samples include additional fine-grained information facilitating the AutoEval task. On three existing datasets and 25 newly introduced ones, we experimentally show that the proposed representation achieves competitive results. Code and dataset are available at https://github.com/sxzrt/Semi-Structured-Dataset-Representations.

Xiaoxiao Sun, Yunzhong Hou, Weijian Deng et al.

Consider a scenario where we are supplied with a number of ready-to-use models trained on a certain source domain and hope to directly apply the most appropriate ones to different target domains based on the models' relative performance. Ideally we should annotate a validation set for model performance assessment on each new target environment, but such annotations are often very expensive. Under this circumstance, we introduce the problem of ranking models in unlabeled new environments. For this problem, we propose to adopt a proxy dataset that 1) is fully labeled and 2) well reflects the true model rankings in a given target environment, and use the performance rankings on the proxy sets as surrogates. We first select labeled datasets as the proxy. Specifically, datasets that are more similar to the unlabeled target domain are found to better preserve the relative performance rankings. Motivated by this, we further propose to search the proxy set by sampling images from various datasets that have similar distributions as the target. We analyze the problem and its solutions on the person re-identification (re-ID) task, for which sufficient datasets are publicly available, and show that a carefully constructed proxy set effectively captures relative performance ranking in new environments. Code is available at \url{https://github.com/sxzrt/Proxy-Set}.

Wenjin Hou, Xiaoxiao Sun, Hehe Fan

Recent advances in zero-shot learning (ZSL) have demonstrated the potential of generative models. Typically, generative ZSL synthesizes visual features conditioned on semantic prototypes to model the data distribution of unseen classes, followed by training a classifier on the synthesized data. However, the synthesized features often remain task-agnostic, leading to degraded performance. Moreover, inferring a faithful distribution from semantic prototypes alone is insufficient for classes that are semantically similar but visually distinct. To address these and advance ZSL, we propose RLVC, an outcome-reward reinforcement learning RL framework with visual cues for generative ZSL. At its core, RL empowers the generative model to self-evolve, implicitly enhancing its generation capability. In particular, RLVC updates the generative model using an outcome-based reward, encouraging the synthesis of task-relevant features. Furthermore, we introduce class-wise visual cues that (i) align synthesized features with visual prototypes and (ii) stabilize the RL training updates. For the training process, we present a novel cold-start strategy. Comprehensive experiments and analyses on three prevalent ZSL benchmarks demonstrate that RLVC achieves state-of-the-art results with a 4.7% gain.

Haoran Lu, Luyang Fang, Ruidong Zhang et al.

Due to the remarkable capabilities and growing impact of large language models (LLMs), they have been deeply integrated into many aspects of society. Thus, ensuring their alignment with human values and intentions has emerged as a critical challenge. This survey provides a comprehensive overview of practical alignment techniques, training protocols, and empirical findings in LLM alignment. We analyze the development of alignment methods across diverse paradigms, characterizing the fundamental trade-offs between core alignment objectives. Our analysis shows that while supervised fine-tuning enables basic instruction-following, preference-based methods offer more flexibility for aligning with nuanced human intent. We discuss state-of-the-art techniques, including Direct Preference Optimization (DPO), Constitutional AI, brain-inspired methods, and alignment uncertainty quantification (AUQ), highlighting their approaches to balancing quality and efficiency. We review existing evaluation frameworks and benchmarking datasets, emphasizing limitations such as reward misspecification, distributional robustness, and scalable oversight. We summarize strategies adopted by leading AI labs to illustrate the current state of practice. We conclude by outlining open problems in oversight, value pluralism, robustness, and continuous alignment. This survey aims to inform both researchers and practitioners navigating the evolving landscape of LLM alignment.

Yinuo Qin, Richard T. Lee, Weijia Zhang et al.

In collaborative environments, a deep understanding of multi-human teaming dynamics is essential for optimizing performance. However, the relationship between individuals' behavioral and physiological markers and their combined influence on overall team performance remains poorly understood. To explore this, we designed a triadic human collaborative sensorimotor task in virtual reality (VR) and introduced a novel predictability metric to examine team dynamics and performance. Our findings reveal a strong connection between team performance and the predictability of a team member's future actions based on other team members' behavioral and physiological data. Contrary to conventional wisdom that high-performing teams are highly synchronized, our results suggest that physiological and behavioral synchronizations among team members have a limited correlation with team performance. These insights provide a new quantitative framework for understanding multi-human teaming, paving the way for deeper insights into team dynamics and performance.

Kartik Patwari, David Schneider, Xiaoxiao Sun et al.

Growing privacy concerns and regulations like GDPR and CCPA necessitate pseudonymization techniques that protect identity in image datasets. However, retaining utility is also essential. Traditional methods like masking and blurring degrade quality and obscure critical context, especially in human-centric images. We introduce Rendering-Refined Stable Diffusion (RefSD), a pipeline that combines 3D-rendering with Stable Diffusion, enabling prompt-based control over human attributes while preserving posture. Unlike standard diffusion models that fail to retain posture or GANs that lack realism and flexible attribute control, RefSD balances posture preservation, realism, and customization. We also propose HumanGenAI, a framework for human perception and utility evaluation. Human perception assessments reveal attribute-specific strengths and weaknesses of RefSD. Our utility experiments show that models trained on RefSD pseudonymized data outperform those trained on real data in detection tasks, with further performance gains when combining RefSD with real data. For classification tasks, we consistently observe performance improvements when using RefSD data with real data, confirming the utility of our pseudonymized data.

Jingyi Zhang, Xiaoxiao Sun

The divide-and-conquer method has been widely used for estimating large-scale kernel ridge regression estimates. Unfortunately, when the response variable is highly skewed, the divide-and-conquer kernel ridge regression (dacKRR) may overlook the underrepresented region and result in unacceptable results. We combine a novel response-adaptive partition strategy with the oversampling technique synergistically to overcome the limitation. Through the proposed novel algorithm, we allocate some carefully identified informative observations to multiple nodes (local processors). Although the oversampling technique has been widely used for addressing discrete label skewness, extending it to the dacKRR setting is nontrivial. We provide both theoretical and practical guidance on how to effectively over-sample the observations under the dacKRR setting. Furthermore, we show the proposed estimate has a smaller risk than that of the classical dacKRR estimate under mild conditions. Our theoretical findings are supported by both simulated and real-data analyses.

Xiaoxiao Sun, Shaomin Mu, Yongyu Xu et al.

In order to identify and prevent tea leaf diseases effectively, convolution neural network (CNN) was used to realize the image recognition of tea disease leaves. Firstly, image segmentation and data enhancement are used to preprocess the images, and then these images were input into the network for training. Secondly, to reach a higher recognition accuracy of CNN, the learning rate and iteration numbers were adjusted frequently and the dropout was added properly in the case of over-fitting. Finally, the experimental results show that the recognition accuracy of CNN is 93.75%, while the accuracy of SVM and BP neural network is 89.36% and 87.69% respectively. Therefore, the recognition algorithm based on CNN is better in classification and can improve the recognition efficiency of tea leaf diseases effectively.

Xiaoxiao Sun, Liang Zheng, Yu-Kun Lai et al.

Annotating a large number of training images is very time-consuming. In this background, this paper focuses on learning from easy-to-acquire web data and utilizes the learned model for fine-grained image classification in labeled datasets. Currently, the performance gain from training with web data is incremental, like a common saying "better than nothing, but not by much". Conventionally, the community looks to correcting the noisy web labels to select informative samples. In this work, we first systematically study the built-in gap between the web and standard datasets, i.e. different data distributions between the two kinds of data. Then, in addition to using web labels, we present an unsupervised object localization method, which provides critical insights into the object density and scale in web images. Specifically, we design two constraints on web data to substantially reduce the difference of data distributions for the web and standard datasets. First, we present a method to control the scale, localization and number of objects in the detected region. Second, we propose to select the regions containing objects that are consistent with the web tag. Based on the two constraints, we are able to process web images to reduce the gap, and the processed web data is used to better assist the standard dataset to train CNNs. Experiments on several fine-grained image classification datasets confirm that our method performs favorably against the state-of-the-art methods.

Xiaoxiao Sun, Liang Zheng

Variations in visual factors such as viewpoint, pose, illumination and background, are usually viewed as important challenges in person re-identification (re-ID). In spite of acknowledging these factors to be influential, quantitative studies on how they affect a re-ID system are still lacking. To derive insights in this scientific campaign, this paper makes an early attempt in studying a particular factor, viewpoint. We narrow the viewpoint problem down to the pedestrian rotation angle to obtain focused conclusions. In this regard, this paper makes two contributions to the community. First, we introduce a large-scale synthetic data engine, PersonX. Composed of hand-crafted 3D person models, the salient characteristic of this engine is "controllable". That is, we are able to synthesize pedestrians by setting the visual variables to arbitrary values. Second, on the 3D data engine, we quantitatively analyze the influence of pedestrian rotation angle on re-ID accuracy. Comprehensively, the person rotation angles are precisely customized from 0 to 360, allowing us to investigate its effect on the training, query, and gallery sets. Extensive experiment helps us have a deeper understanding of the fundamental problems in person re-ID. Our research also provides useful insights for dataset building and future practical usage, e.g., a person of a side view makes a better query.