Author name cluster

Xiaofeng Zhang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

25 papers

2 author rows

EAAI Journal 2026 Journal Article

A pretrained self-supervised model with lightweight spatial feature distillation for wetland mapping via satellite imagery

Bingqian Wang
Jianhua Chen
Huajun Wang
Shixiang Zuo
Xiaofeng Zhang
Qian Zhang
Yipeng Tang
Jiongling Chen

Wetlands are globally important ecosystems owing to their unique resources and environmental advantages. However, indiscriminate development and utilization have led to significant wetland resource loss. Employing remote sensing technology for long-term wetland mapping is important for wetland conservation and sustainable utilization. The demand for long-term spans makes the remote sensing data used in this study stringent, and it also renders the labeling of many samples, impractical. To address these issues, we propose a model called the Pretrained Self-supervised visual foundation model with Lightweight Spatial Feature Distillation (PSLSFD). This model adopts transfer learning by extracting deep features using a frozen pretrained backbone and trains only a lightweight distillation head to distill these pretrained features from both spatial and spectral perspectives with the aim of extracting compact within-class and discriminative between-class structures, which significantly reduces training cost compared with fine-tuning the backbone, and the inference cost is dominated by the frozen backbone. Finally, we applied the model to wetland mapping in Ruoergai National Park. The results of the comparative experiments indicated that the overall classification accuracy of PSLSFD using linear probes reached 98. 2%, whereas that using clustering probes was 60. 5%. These experimental results outperformed those of other wetland landscape mapping algorithms. Additionally, by creating a wetland landscape classification dataset covering nearly 30 years for Ruoergai National Park, we demonstrated that PSLSFD can be applied to subsequent wetland evolution studies.

Details DOI

AAAI Conference 2026 Conference Paper

D3ToM: Decider-Guided Dynamic Token Merging for Accelerating Diffusion MLLMs

Shuochen Chang
Xiaofeng Zhang
Qingyang Liu
Li Niu

Diffusion-based multimodal large language models (Diffusion MLLMs) have recently demonstrated impressive non-autoregressive generative capabilities across vision-and-language tasks. However, Diffusion MLLMs exhibit substantially slower inference than autoregressive models: Each denoising step employs full bidirectional self-attention over the entire sequence, resulting in cubic decoding complexity that becomes computationally impractical with thousands of visual tokens. To address this challenge, we propose D³ToM, a Decider-guided dynamic token merging method that dynamically merges redundant visual tokens at different denoising steps to accelerate inference in Diffusion MLLMs. At each denoising step, D³ToM uses decider tokens—the tokens generated in the previous denoising step—to build an importance map over all visual tokens. Then it maintains a proportion of the most salient tokens and merges the remainder through similarity-based aggregation. This plug-and-play module integrates into a single transformer layer, physically shortening the visual token sequence for all subsequent layers without altering model parameters. Moreover, D³ToM employs a merge ratio that dynamically varies with each denoising step, aligns with the native decoding process of Diffusion MLLMs, achieving superior performance under equivalent computational budgets. Extensive experiments show that D³ToM accelerates inference while preserving competitive performance.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Diversity Recommendation via Causal Deconfounding of Co-purchase Relations and Counterfactual Exposure

Jingmao Zhang
Zhiting Zhao
Yunqi Lin
Jianghong Ma
Tianjun Wei
Haijun Zhang
Xiaofeng Zhang

Beyond user-item modeling, item-to-item relationships are increasingly used to enhance recommendation. However, common methods largely rely on co-occurrence, making them prone to item popularity bias and user attributes, which degrades embedding quality and performance. Meanwhile, although diversity is acknowledged as a key aspect of recommendation quality, existing research offers limited attention to it, with a notable lack of causal perspectives and theoretical grounding. To address these challenges, we propose Cadence: Diversity Recommendation via Causal Deconfounding of Co-purchase Relations and Counterfactual Exposure—a plug-and-play framework built upon LightGCN as the backbone, primarily designed to enhance recommendation diversity while preserving accuracy. First, we compute the Unbiased Asymmetric Co-purchase Relationship (UACR) between items—excluding item popularity and user attributes—to construct a deconfounded directed item graph, with an aggregation mechanism to refine embeddings. Second, we leverage UACR to identify diverse categories of items that exhibit strong causal relevance to a user's interacted items but have not yet been engaged with. We then simulate their behavior under high-exposure scenarios, thereby significantly enhancing recommendation diversity while preserving relevance. Extensive experiments on real-world datasets demonstrate that our method consistently outperforms state-of-the-art diversity models in both diversity and accuracy, and further validates its effectiveness, transferability, and efficiency over baselines.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Remember Me: Bridging the Long-Range Gap in LVLMs with Three-Step Inference-Only Decay Resilience Strategies

Peng Gao
Yujian Lee
Xiaofeng Zhang
Zailong Chen
Hui Zhang

Large Vision-Language Models (LVLMs) have achieved impressive performance across a wide range of multimodal tasks. However, they still face critical challenges in modeling long-range dependencies under the usage of Rotary Positional Encoding (ROPE). Although it can facilitate precise modeling of token positions, it induces progressive attention decay as token distance increases, especially with progressive attention decay over distant token pairs, which severely impairs the model's ability to remember global context. To alleviate this issue, we propose inference-only Three-step Decay Resilience Strategies (T-DRS), comprising (1) Semantic-Driven DRS (SD-DRS), amplifying semantically meaningful but distant signals via content-aware residuals, (2) Distance-aware Control DRS (DC-DRS), which can purify attention by smoothly modulating weights based on positional distances, suppressing noise while preserving locality, and (3) re-Reinforce Distant DRS (reRD-DRS), consolidating the remaining informative remote dependencies to maintain global coherence. Together, the T-DRS recover suppressed long-range token pairs without harming local inductive biases. Extensive experiments on Vision Question Answering (VQA) benchmarks demonstrate that T-DRS can consistently improve performance in an inference-only manner.

PDF Details DOI

AAAI Conference 2026 Conference Paper

SGMT: Social Generating with Multiview-Guided Tuning In Recommender Systems

Jianghong Ma
Changran He
Dezhao Yang
Tianjun Wei
Haijun Zhang
Xiaofeng Zhang

The sparsity of user–item interactions remains a fundamental obstacle in collaborative filtering, limiting the ability of Graph Neural Network (GNN)-based recommender systems to capture high-order user relationships without incurring over-smoothing and computational overhead. Existing social recommendation approaches mitigate this by incorporating social networks, yet most rely on explicit ties and fail to construct informative links in their absence. Meanwhile, contrastive learning (CL) has shown promise in improving representation quality, but current view generation strategies, augmentation-based for robustness and nonaugmentation-based for semantic fidelity, are seldom combined, leaving their complementary potential underexplored. We propose Social Generating with Multiview-guided Tuning (SGMT), a unified framework that addresses both challenges. First, an interest-aware social generation mechanism constructs synthetic user–user links from shared interaction patterns, theoretically shown to compress collaborative paths and uncover latent high-order relations. Second, we present two complementary CL modules, Noise-augmented View and Semantic-explored View, which we theoretically prove to preferentially enhance uniformity and alignment, respectively, two fundamental objectives in CL. Experiments on three real-world datasets show that SGMT outperforms state-of-the-art baselines, validating both the theoretical analysis and the practical efficacy of our model.

PDF Details DOI

EAAI Journal 2025 Journal Article

A novel dual-channel model with adaptive multi-scale attention for time series forecasting

Shuqing Wang
Jinghao Lu
Ren Wang
Xiaofeng Zhang
Hua Wang
Yujuan Sun

Time series forecasting plays a crucial role in various domains, including finance, traffic management, energy, and healthcare. However, as application scenarios continue to expand, the complexity of time series data has significantly increased, posing substantial challenges in capturing trend fluctuations of multivariate features and the dynamic relationships among them. To address these issues, this paper proposes a novel architecture–DASformer (Dual-Channel model with Adaptive multi-Scale attention) - which enhances time series analysis by leveraging a dual-channel multivariate extractor and an adaptive multi-scale attention mechanism. Specifically, the dual-channel multivariate extractor comprises two independent yet interactive streams, focusing on capturing information at different levels of the time series, thereby effectively decoupling complex dynamic relationships. Moreover, to alleviate the problem of feature forgetting and loss in the long-term trend stream, the model incorporates an adaptive multi-scale attention module. This module adopts multi-scale processing and a dynamic weighting mechanism to learn dependencies across different scales and effectively capture their dynamic variations. Experimental results show that DASformer consistently achieves state-of-the-art performance on nine widely used benchmark datasets, delivering superior prediction accuracy, particularly in long-term forecasting tasks. The source code is available at: https: //github. com/LDU-TSA/DASformer.

Details DOI

IROS Conference 2025 Conference Paper

AdaToken-3D: Dynamic Spatial Gating for Efficient 3D Large Multimodal-Models Reasoning

Kai Zhang
Xingyu Chen
Xiaofeng Zhang

Large Multimodal Models (LMMs) have become a pivotal research focus in deep learning, demonstrating remarkable capabilities in 3D scene understanding. However, current 3D LMMs employing thousands of spatial tokens for multimodal reasoning suffer from critical inefficiencies: excessive computational overhead and redundant information flows. Unlike 2D VLMs processing single images, 3D LMMs exhibit inherent architectural redundancy due to the heterogeneous mechanisms between spatial tokens and visual tokens. To address this challenge, we propose AdaToken-3D, an adaptive spatial token optimization framework that dynamically prunes redundant tokens through spatial contribution analysis. Our method automatically tailors pruning strategies to different 3D LMM architectures by quantifying token-level information flows via attention pattern mining. Extensive experiments on LLaVA-3D (a 7B parameter 3D-LMM) demonstrate that AdaToken-3D achieves 21% faster inference speed and 63% FLOPs reduction while maintaining original task accuracy. Beyond efficiency gains, this work systematically investigates redundancy patterns in multimodal spatial information flows through quantitative token interaction analysis. Our findings reveal that over 60% of spatial tokens contribute minimally (<5%) to the final predictions, establishing theoretical foundations for efficient 3D multimodal learning.

Details

TIST Journal 2025 Journal Article

An Underwater Imaging Generative Adversarial Network by Simulating the Mechanism of Light Propagation in Water

Yujuan Sun
Xing Huang
Yanfang Cui
Junyu Dong
Xiaofeng Zhang
Tao Yao

Since capturing underwater images without degradation is challenging, there are few real image datasets with paired ground truth for underwater image enhancement. In this article, we propose a generative adversarial network (UIGAN) for underwater imaging; the network can convert images and their corresponding depth maps captured in air into images in water. The underwater imaging mechanism relies on many intrinsic parameters in water, which are difficult to estimate without field calibration. As the strong modeling capability of deep neural networks, this article uses the deep learning model to extract parameters from the real underwater environment. Then the proposed UIGAN simulates the light propagation process (direct attenuation, backscattering, and forward scattering) in water by using three modules with different constraints. We can generate a large training dataset with paired images in air and real water environment. The generated UIGAN dataset serves as input to a forward-attention transfer underwater enhancement model (FATUECNN), and it can output the restored images with appearance like those captured in air. The proposed pipeline is verified both qualitatively and quantitatively by extensive experiments and comparison evaluation with the existing state-of-the-art methods. The source code and the pre-trained model are made publicly available.

Details DOI

AAAI Conference 2025 Conference Paper

CoDeR: Counterfactual Demand Reasoning for Sequential Recommendation

Shuai Tang
Sitao Lin
Jianghong Ma
Xiaofeng Zhang

Sequential recommendation systems aim to predict the next item based on users' historical interactions. While traditional methods focus on learning feature representations or user preferences, they often struggle with detecting subtle demand shifts in short sequences, especially when these shifts are obscured by noise or biases. To address these issues, we propose CoDeR (Counterfactual Demand Reasoning), a novel framework designed to handle demand shifts in sequential recommendations with greater precision. CoDeR features two key modules: (1) the User Demand Extraction module, which utilizes self-attention mechanisms and demand graphs to identify and model demand shifts from minimal user interactions; and (2) the Counterfactual Demand Reasoning module, which employs causal effect analysis and backdoor adjustment techniques to distinguish true demand shifts from noisy or biased signals. Our approach represents the first application of counterfactual reasoning to sequential recommendation systems. Comprehensive experiments on three real-world datasets demonstrate that CoDeR significantly outperforms existing baselines.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

CURV: Coherent Uncertainty-Aware Reasoning in Vision-Language Models for X-Ray Report Generation

Ziao Wang
Sixing Yan
Kejing Yin
Xiaofeng Zhang
William K. Cheung

Vision-language models have been explored for radiology report generation with promising results. Yet, uncertainty elaborated in findings and the reasoning process for reaching clinical impressions are seldom explicitly modeled, reducing the clinical accuracy and trustworthiness of the generated reports. We present CURV, a novel framework that alleviates the limitations through integrated awareness of uncertainty and explicit reasoning capabilities. Our approach consists of three key components: (1) an uncertainty modeling mechanism that teaches the model to recognize and express appropriate levels of diagnostic confidence, (2) a structured reasoning framework that generates intermediate explanatory steps connecting visual findings to clinical impressions, and (3) a reasoning coherence reward that ensures logical consistency among findings, reasoning, and impressions. We implement CURV through a three-stage training pipeline that combines uncertainty-aware fine-tuning, reasoning initialization, and reinforcement learning. In particular, we adopt a comprehensive reward function addresses multiple aspects of report quality, incorporating medical term matching, uncertainty expression evaluation, and semantic coherence evaluation. Experimental results demonstrate that CURV generates clinically relevant reports with appropriate uncertainty expressions and transparent reasoning traces, significantly outperforming previous methods. CURV represents a substantial advancement toward interpretable and trustworthy AI-generated radiology reports, with broader implications for the deployment of vision-language models in high-stakes clinical environments where uncertainty awareness and reasoning transparency are essential.

PDF Details

EAAI Journal 2025 Journal Article

Enhanced pediatric pneumonia auxiliary diagnosis: Integrating optical fiber vibration sensing with machine learning

Pengfei Cao
Jiawei Xu
Yifan Zhao
Qian Ni
Yuxia Li
Hansen Chen
Ming Song
Jiqiang Shang

Childhood pneumonia remains a primary cause of death in children under five. Early detection is arduous due to its inconspicuous symptoms, and the existing radiological diagnostic techniques carry the risk of radiation - induced harm to young patients. To overcome these limitations, a groundbreaking study has proposed an innovative non - invasive diagnostic approach that integrates fiber optic vibration sensing technology with machine learning algorithms for pediatric pneumonia diagnosis. A novel fiber optic sensor is engineered to precisely capture respiratory vibration signals (RVS). These signals are then processed and analyzed using a Stacked - Grid Search Ensemble Learning Model (SGELM). In the experiment, respiratory vibration signals were gathered from 1649 pediatric patients aged between 3 and 14 who suffered from respiratory diseases. Through data balancing techniques, the dataset was expanded to 2184 samples. This dataset was partitioned into training, testing, and validation subsets. The developed system exhibited remarkable performance on the test dataset. It achieved high levels of accuracy, sensitivity, and specificity. Notably, it was also capable of classifying different pneumonia pathological types and statuses. This innovative method not only mitigates the radiation - related risks associated with traditional diagnostic methods but also holds great promise in revolutionizing the diagnosis of pediatric respiratory diseases. It could potentially improve the early - diagnosis rate and contribute to better treatment outcomes for children with pneumonia, thus playing a significant role in enhancing child health globally.

Details DOI

AAAI Conference 2025 Conference Paper

Enhancing Multimodal Large Language Models Complex Reason via Similarity Computation

Xiaofeng Zhang
Fanshuo Zeng
Yihao Quan
Zheng Hui
Jiawei Yao

Multimodal large language models have experienced rapid growth, and numerous different models have emerged. The interpretability of LVLMs remains an under-explored area. Especially when faced with more complex tasks such as chain-of-thought reasoning, its internal mechanisms still resemble a black box that is difficult to decipher. By studying the interaction and information flow between images and text, we noticed that in models such as LLaVA1.5, image tokens that are semantically related to text are more likely to have information flow convergence in the LLM decoding layer, and these image tokens receive higher attention scores. However, those image tokens that are less relevant to the text do not have information flow convergence, and they only get very small attention scores. To efficiently utilize the image information, we propose a new image token reduction method, Simignore, which aims to improve the complex reasoning ability of LVLMs by computing the similarity between image and text embeddings and ignoring image tokens that are irrelevant and unimportant to the text. Through extensive experiments, we demonstrate the effectiveness of our method for complex reasoning tasks.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Fine-Grained Preference Optimization Improves Spatial Reasoning in VLMs

Yifan Shen
Yuanzhe Liu
Jingyuan Zhu
Xu Cao
Xiaofeng Zhang
Yixiao He
Wenming Ye
James Rehg

Current Vision-Language Models (VLMs) struggle with fine-grained spatial reasoning, particularly when multi-step logic and precise spatial alignment are required. In this work, we introduce SpatialReasoner-R1, a vision-language reasoning model designed to address these limitations. To construct high-quality supervision for spatial reasoning, we design a Multi-Model Monte Carlo Tree Search (M3CTS) method that generates diverse, logically consistent Long Chain-of-Thought (LongCoT) reasoning trajectories. In addition, we propose a fine-grained Direct Preference Optimization (fDPO) method that introduces segment-specific preference granularity for descriptive grounding and logical reasoning, guided by a spatial reward mechanism that evaluates candidate responses based on visual consistency, spatial grounding, and logical coherence. Experimental results demonstrate that fDPO achieves relative performance gains of 4. 1% and 9. 0% over standard DPO on spatial qualitative and quantitative tasks, respectively. SpatialReasoner-R1, trained with fDPO, sets a new SoTA on SpatialRGPT-Bench, outperforming the strongest baseline by 9. 4% in average accuracy, while maintaining competitive performance on general vision-language tasks.

PDF Details

NeurIPS Conference 2025 Conference Paper

Increasing the Utility of Synthetic Images through Chamfer Guidance

Nicola DallAsen
Xiaofeng Zhang
Reyhane Askari Hemmat
Melissa Hall
Jakob Verbeek
Adriana Romero-Soriano
Michal Drozdzal

Conditional image generative models hold considerable promise to produce infinite amounts of synthetic training data. Yet, recent progress in generation quality has come at the expense of generation diversity, limiting the utility of these models as a source of synthetic training data. Although guidance-based approaches have been introduced to improve the utility of generated data by focusing on quality or diversity, the (implicit or explicit) utility functions oftentimes disregard the potential distribution shift between synthetic and real data. In this work, we introduce Chamfer Guidance: a training-free guidance approach which leverages a handful of real exemplar images to characterize the quality and diversity of synthetic data. We show that by leveraging the proposed Chamfer Guidance, we can boost the diversity of the generations w. r. t. a dataset of real images while maintaining or improving the generation quality on ImageNet-1k and standard geo-diversity benchmarks. Our approach achieves state-of-the-art few-shot performance with as little as 2 exemplar real images, obtaining 96. 4% in terms of precision, and 86. 4% in terms of distributional coverage, which increase to 97. 5% and 92. 7%, respectively, when using 32 real images. We showcase the benefits of the Chamfer Guidance generation by training downstream image classifiers on synthetic data, achieving accuracy boost of up to 15% for in-distribution over the baselines, and up to 16\% in out-of-distribution. Furthermore, our approach does not require using the unconditional model, and thus obtains a 31\% reduction in FLOPs w. r. t. classifier-free-guidance-based approaches at sampling time.

PDF Details

IJCAI Conference 2025 Conference Paper

LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction

Jiesong Bai
Yuhao Yin
Yihang Dong
Xiaofeng Zhang
Chi-Man Pun
Xuhang Chen

Lensless imaging stands out as a promising alternative to conventional lens-based systems, particularly in scenarios demanding ultracompact form factors and cost-effective architectures. However, such systems are fundamentally governed by the Point Spread Function (PSF), which dictates how a point source contributes to the final captured signal. Traditional lensless techniques often require explicit calibrations and extensive pre-processing, relying on static or approximate PSF models. These rigid strategies can result in limited adaptability to real-world challenges, including noise, system imperfections, and dynamic scene variations, thus impeding high-fidelity reconstruction. In this paper, we propose LensNet, an end-to-end deep learning framework that integrates spatial-domain and frequency-domain representations in a unified pipeline. Central to our approach is a learnable Coded Mask Simulator (CMS) that enables dynamic, data-driven estimation of the PSF during training, effectively mitigating the shortcomings of fixed or sparsely calibrated kernels. By embedding a Wiener filtering component, LensNet refines global structure and restores fine-scale details, thus alleviating the dependency on multiple handcrafted pre-processing steps. Extensive experiments demonstrate LensNet's robust performance and superior reconstruction quality compared to state-of-the-art methods, particularly in preserving high-frequency details and attenuating noise. The proposed framework establishes a novel convergence between physics-based modeling and data-driven learning, paving the way for more accurate, flexible, and practical lensless imaging solutions for applications ranging from miniature sensors to medical diagnostics. The link of code is https: //github. com/baijiesong/Lensnet.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Beyond What If: Advancing Counterfactual Text Generation with Structural Causal Modeling

Ziao Wang
Xiaofeng Zhang
Hongwei Du

Exploring the realms of counterfactuals, this paper introduces a versatile approach in text generation using structural causal models (SCM), broadening the scope beyond traditional singular causal studies to encompass complex, multi-layered relationships. To comprehensively explore these intricate, multi-layered causal relationships in text generation, we introduce a generalized approach based on the structural causal model (SCM), adept at handling complex causal interactions in a spectrum ranging from everyday stories to financial reports. Specifically, our method begins by disentangling each component of the text into pairs of latent variables, representing elements that remain unchanged and those subject to variation. Subsequently, counterfactual interventions are applied to these latent variables, facilitating the generation of outcomes that are influenced by complex causal dynamics. Extensive experiments have been conducted on both a public story generation dataset and a specially constructed dataset in the financial domain. The experimental results demonstrate that our approach achieves state-of-the-art performance across a range of automatic and human evaluation criteria, underscoring its effectiveness and versatility in diverse text generation contexts.

PDF Details DOI

YNIMG Journal 2024 Journal Article

Brain extended and closed forms glutathione levels decrease with age and extended glutathione is associated with visuospatial memory

Xin Hu
Keyu Pan
Min Zhao
Jiali Lv
Jing Wang
Xiaofeng Zhang
Yuxi Liu
Yulu Song

During aging, the brain is subject to greater oxidative stress (OS), which is thought to play a critical role in cognitive impairment. Glutathione (GSH), as a major antioxidant in the brain, can be used to combat OS. However, how brain GSH levels vary with age and their associations with cognitive function is unclear. In this study, we combined point-resolved spectroscopy and edited spectroscopy sequences to investigate extended and closed forms GSH levels in the anterior cingulate cortex (ACC), posterior cingulate cortex (PCC), and occipital cortex (OC) of 276 healthy participants (extended form, 166 females, age range 20-70 years) and 15 healthy participants (closed form, 7 females, age range 26-56 years), and examined their relationships with age and cognitive function. The results revealed decreased extended form GSH levels with age in the PCC among 276 participants. Notably, the timecourse of extended form GSH level changes in the PCC and ACC differed between males and females. Additionally, positive correlations were observed between extended form GSH levels in the PCC and OC and visuospatial memory. Additionally, a decreased trend of closed form GSH levels with age was also observed in the PCC among 15 participants. Taken together, these findings enhance our understanding of the brain both closed and extended form GSH time course during normal aging and associations with sex and memory, which is an essential first step for understanding the neurochemical underpinnings of healthy aging.

Details DOI

YNIMG Journal 2024 Journal Article

Deep learning enables accurate brain tissue microstructure analysis based on clinically feasible diffusion magnetic resonance imaging

Yuxing Li
Zhizheng Zhuo
Chenghao Liu
Yunyun Duan
Yulu Shi
Tingting Wang
Runzhi Li
Yanli Wang

Diffusion magnetic resonance imaging (dMRI) allows non-invasive assessment of brain tissue microstructure. Current model-based tissue microstructure reconstruction techniques require a large number of diffusion gradients, which is not clinically feasible due to imaging time constraints, and this has limited the use of tissue microstructure information in clinical settings. Recently, approaches based on deep learning (DL) have achieved promising tissue microstructure reconstruction results using clinically feasible dMRI. However, it remains unclear whether the subtle tissue changes associated with disease or age are properly preserved with DL approaches and whether DL reconstruction results can benefit clinical applications. Here, we provide the first evidence that DL approaches to tissue microstructure reconstruction yield reliable brain tissue microstructure analysis based on clinically feasible dMRI scans. Specifically, we reconstructed tissue microstructure from four different brain dMRI datasets with only 12 diffusion gradients, a clinically feasible protocol, and the neurite orientation dispersion and density imaging (NODDI) and spherical mean technique (SMT) models were considered. With these results we show that disease-related and age-dependent alterations of brain tissue were accurately identified. These findings demonstrate that DL tissue microstructure reconstruction can accurately quantify microstructural alterations in the brain based on clinically feasible dMRI.

Details DOI

NeurIPS Conference 2024 Conference Paper

Instance-adaptive Zero-shot Chain-of-Thought Prompting

Xiaosong Yuan
Chen Shen
Shaotian Yan
Xiaofeng Zhang
Liang Xie
Wenxiao Wang
Renchu Guan
Ying Wang

Zero-shot Chain-of-Thought (CoT) prompting emerges as a simple and effective strategy for enhancing the performance of large language models (LLMs) in real-world reasoning tasks. Nonetheless, the efficacy of a singular, task-level prompt uniformly applied across the whole of instances is inherently limited since one prompt cannot be a good partner for all, a more appropriate approach should consider the interaction between the prompt and each instance meticulously. This work introduces an instance-adaptive prompting algorithm as an alternative zero-shot CoT reasoning scheme by adaptively differentiating good and bad prompts. Concretely, we first employ analysis on LLMs through the lens of information flow to detect the mechanism under zero-shot CoT reasoning, in which we discover that information flows from question to prompt and question to rationale jointly influence the reasoning results most. We notice that a better zero-shot CoT reasoning needs the prompt to obtain semantic information from the question then the rationale aggregates sufficient information from the question directly and via the prompt indirectly. On the contrary, lacking any of those would probably lead to a bad one. Stem from that, we further propose an instance-adaptive prompting strategy (IAP) for zero-shot CoT reasoning. Experiments conducted with LLaMA-2, LLaMA-3, and Qwen on math, logic, and commonsense reasoning tasks (e. g. , GSM8K, MMLU, Causal Judgement) obtain consistent improvement, demonstrating that the instance-adaptive zero-shot CoT prompting performs better than other task-level methods with some curated prompts or sophisticated procedures, showing the significance of our findings in the zero-shot CoT reasoning mechanism.

PDF Details DOI

YNIMG Journal 2024 Journal Article

SiMix: A domain generalization method for cross-site brain MRI harmonization via site mixing

Chundan Xu
Jie Li
Yakui Wang
Lixue Wang
Yizhe Wang
Xiaofeng Zhang
Weiqi Liu
Jingang Chen

Brain magnetic resonance imaging (MRI) is widely used in clinical practice for disease diagnosis. However, MRI scans acquired at different sites can have different appearances due to the difference in the hardware, pulse sequence, and imaging parameter. It is important to reduce or eliminate such cross-site variations with brain MRI harmonization so that downstream image processing and analysis is performed consistently. Previous works on the harmonization problem require the data acquired from the sites of interest for model training. But in real-world scenarios there can be test data from a new site of interest after the model is trained, and training data from the new site is unavailable when the model is trained. In this case, previous methods cannot optimally handle the test data from the new unseen site. To address the problem, in this work we explore domain generalization for brain MRI harmonization and propose Site Mix (SiMix). We assume that images of travelling subjects are acquired at a few existing sites for model training. To allow the training data to better represent the test data from unseen sites, we first propose to mix the training images belonging to different sites stochastically, which substantially increases the diversity of the training data while preserving the authenticity of the mixed training images. Second, at test time, when a test image from an unseen site is given, we propose a multiview strategy that perturbs the test image with preserved authenticity and ensembles the harmonization results of the perturbed images for improved harmonization quality. To validate SiMix, we performed experiments on the publicly available SRPBS dataset and MUSHAC dataset that comprised brain MRI acquired at nine and two different sites, respectively. The results indicate that SiMix improves brain MRI harmonization for unseen sites, and it is also beneficial to the harmonization of existing sites.

Details DOI

IJCAI Conference 2023 Conference Paper

Beyond Pure Text: Summarizing Financial Reports Based on Both Textual and Tabular Data

Ziao Wang
Zelin Jiang
Xiaofeng Zhang
Jaehyeon Soon
Jialu Zhang
Wang Xiaoyao
Hongwei Du

Abstractive text summarization is to generate concise summaries that well preserve both salient information and the overall semantic meanings of the given documents. However, real-world documents, e. g. , financial reports, generally contain rich data such as charts and tabular data which invalidates most existing text summarization approaches. This paper is thus motivated to propose this novel approach to simultaneously summarize both textual and tabular data. Particularly, we first manually construct a “table+text → summary” dataset. Then, the tabular data is respectively embedded in a row-wise and column-wise manner, and the textual data is encoded at the sentence-level via an employed pre-trained model. We propose a salient detector gate respectively performed between each pair of row/column and sentence embeddings. The highly correlated content is considered as salient information that must be summarized. Extensive experiments have been performed on our constructed dataset and the promising results demonstrate the effectiveness of the proposed approach w. r. t. a number of both automatic and human evaluation criteria.

PDF Details DOI

AAAI Conference 2021 Conference Paper

Demodalizing Face Recognition with Synthetic Samples

Zhonghua Zhai
Pengju Yang
Xiaofeng Zhang
Maji Huang
Haijing Cheng
Xuejun Yan
Chunmao Wang
Shiliang Pu

Using data generated by generative adversarial networks or three-dimensional (3D) technology for face recognition training is a theoretically reasonable solution to the problems of unbalanced data distributions and data scarcity. However, due to the modal difference between synthetic data and real data, the direct use of data for training often leads to a decrease in the recognition performance, and the effect of synthetic data on recognition remains ambiguous. In this paper, after observing in experiments that modality information has a fixed form, we propose a demodalizing face recognition training architecture for the first time and provide a feasible method for recognition training using synthetic samples. Specifically, three different demodalizing training methods, from implicit to explicit, are proposed. These methods gradually reveal a generated modality that is difficult to quantify or describe. By removing the modalities of the synthetic data, the performance degradation is greatly alleviated. We validate the effectiveness of our approach on various benchmarks of largescale face recognition and outperform the previous methods, especially in the low FAR range.

PDF Details

AAAI Conference 2021 Short Paper

Generating Long Financial Report using Conditional Variational Autoencoders with Knowledge Distillation

Yunpeng Ren
Ziao Wang
Yiyuan Wang
Xiaofeng Zhang

Automatically generating financial report from a piece of news is quite a challenging task. Apparently, the difficulty of this task lies in the lack of sufficient background knowledge to effectively generate long financial report. To address this issue, this paper proposes the conditional variational autoencoders (CVAE) based approach which distills external knowledge from a corpus of news-report data, and experimental results show that it can achieve the SOTA performance.

PDF Details

YNIMG Journal 2007 Journal Article

A spatial and temporal comparison of hemodynamic signals measured using optical and functional magnetic resonance imaging during activation in the human primary visual cortex

Vladislav Y. Toronov
Xiaofeng Zhang
Andrew G. Webb

Functional near infrared spectro-imaging (fNIRSI) is potentially a very useful technique for obtaining information about the underlying physiology of the blood oxygenation level dependent (BOLD) signal used in functional magnetic resonance imaging (fMRI). In this paper the temporal and spatial statistical characteristics of fNIRSI data are compared to those of simultaneously acquired fMRI data in the human visual cortex during a variable-frequency reversing checkerboard activation paradigm. Changes in the size of activated volume caused by changes in checkerboard reversal frequency allowed a comparison of the behavior of the spatial responses measured by the two imaging methods. fNIRSI and fMRI data were each analyzed using standard correlation and fixed-effect group analyses of variance pathways. The statistical significance of fNIRSI data was found to be much lower than that of the fMRI data, due mainly to the low signal-to-noise of the measurements. Reconstructed images also showed that, while the time-course of changes in the oxy-, deoxy-, and total hemoglobin concentrations all exhibit high correlation with that of the BOLD response, the changes in the volume of tissue measured as “activated” by the BOLD response demonstrate a closer similarity to the corresponding changes in the oxy- and total hemoglobin concentrations than to that of the deoxyhemoglobin.

Details DOI

IJCAI Conference 2005 Conference Paper

Learning Global Models Based on Distributed Data Abstractions

Xiaofeng Zhang
William K

Due to the increasing demand of massive and distributed data analysis, achieving highly accurate global data analysis results with local data privacy preserved becomes an increasingly important research issue. In this paper, we propose to adopt a model-based method (Gaussian mixture model) for local data abstraction and aggregate the local model parameters for learning global models. To support global model learning based on solely local GMM parameters instead of virtual data generated from the aggregated local model, a novel EM-like algorithm is derived. Experiments have been performed using synthetic datasets and the proposed method was demonstrated to be able to achieve the global model accuracy comparable to that of using the data regeneration approach at a much lower computational cost.

PDF