papers

Publications (71)

astro-ph1995

High Resolution IRAS Maps and IR Emission of M31 --- II. Diffuse Component and Interstellar Dust

Cong Xu, George Helou

Large-scale dust heating and cooling in the diffuse medium of M31 is studied using the HiRes IRAS maps in conjunction with UV, optical (UBV) and the HI maps. A dust heating/cooling…

cs.LG2020

Improve Adversarial Robustness via Weight Penalization on Classification Layer

Cong Xu, Dan Li, Min Yang

It is well-known that deep neural networks are vulnerable to adversarial attacks. Recent studies show that well-designed classification parts can lead to better robustness. However…

cs.LG2025

Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection

Yihua Wang, Qi Jia, Cong Xu +6

Multimodal sarcasm detection is a complex task that requires distinguishing subtle complementary signals across modalities while filtering out irrelevant information. Many advanced…

cs.LG2017

TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning

Wei Wen, Cong Xu, Feng Yan +4

High network communication cost for synchronizing gradients and parameters is the well-known bottleneck of distributed training. In this work, we propose TernGrad that uses ternary…

astro-ph1998

Starburst in the Intragroup Medium of Stephan's Quintet

Cong Xu, Richard Tuffs

Based on new ISO mid-infrared observations and ground based and near-infrared observations, we report the detection of a bright starburst in the intragroup medium (IGM) of t…

quant-ph2024

Uncertainty of quantum channels based on symmetrized \r{ho}-absolute variance and modified Wigner-Yanase skew information

Cong Xu, Qing-Hua Zhang, Shao-Ming Fei

We present the uncertainty relations in terms of the symmetrized \r{ho}-absolute variance, which generalizes the uncertainty relations for arbitrary operator (not necessarily Hermi…

stat.AP2026

A Utility Score Framework for Dose Optimization Studies with Binary Efficacy-Safety Endpoints: Sample Size Determination and Bias Characterization

Xuemin Gu, Cong Xu, Lei Xu +1

The FDA's Project Optimus initiative emphasizes patient-centered dose selection in oncology that balances efficacy and safety. We develop a framework for randomized dose optimizati…

cs.LG2018

SmoothOut: Smoothing Out Sharp Minima to Improve Generalization in Deep Learning

Wei Wen, Yandan Wang, Feng Yan +4

In Deep Learning, Stochastic Gradient Descent (SGD) is usually selected as a training method because of its efficiency; however, recently, a problem in SGD gains research interest:…

quant-ph2022

Tighter uncertainty relations based on modified weighted Wigner-Yanase-Dyson skew information of quantum channels

Cong Xu, Zhaoqi Wu, Shao-Ming Fei

We use a novel formation to illustrate the () modified weighted Wigner-Yanase-Dyson (() MWWYD) skew information of quantum channels. By using operator norm ineq…

cs.LG2023

Missingness Augmentation: A General Approach for Improving Generative Imputation Models

Yufeng Wang, Dan Li, Cong Xu +1

Missing data imputation is a fundamental problem in data analysis, and many studies have been conducted to improve its performance by exploring model structures and learning proced…

astro-ph2002

Far-Infrared photometry of a statistical sampleof late-type Virgo Cluster galaxies

Richard J. Tuffs, Cristina C. Popescu, Daniele Pierini +6

We present deep diffraction-limited far-infrared (FIR) strip maps of a sample of 63 galaxies later than S0 and brighter than B_T 16.8, selected from the Virgo Cluster Catalogue of…

cs.LG2025

X-TIME: An in-memory engine for accelerating machine learning on tabular data with CAMs

Giacomo Pedretti, John Moon, Pedro Bruel +11

Structured, or tabular, data is the most common format in data science. While deep learning models have proven formidable in learning from unstructured data such as images or speec…

cs.IR2024

Are LLM-based Recommenders Already the Best? Simple Scaled Cross-entropy Unleashes the Potential of Traditional Sequential Recommenders

Cong Xu, Zhangchi Zhu, Mo Yu +3

Large language models (LLMs) have been garnering increasing attention in the recommendation community. Some studies have observed that LLMs, when fine-tuned by the cross-entropy (C…

astro-ph.IM2021

Change point detection and image segmentation for time series of astrophysical images

Cong Xu, Hans Moritz Günther, Vinay L. Kashyap +2

Many astrophysical phenomena are time-varying, in the sense that their intensity, energy spectrum, and/or the spatial distribution of the emission suddenly change. This paper devel…

cs.CV2022

Improve Deep Image Inpainting by Emphasizing the Complexity of Missing Regions

Yufeng Wang, Dan Li, Cong Xu +1

Deep image inpainting research mainly focuses on constructing various neural network architectures or imposing novel optimization objectives. However, on the one hand, building a s…

gr-qc2026

Probing Gravitational Quantum Field Theory through Polarization Fingerprints of Gravitational Waves

Cong Xu, Hong-Bo Jin, Yue-Liang Wu

Gravitational Quantum Field Theory (GQFT) has been proposed as a candidate framework to reconcile general relativity with quantum field theory, and a distinctive imprint on gravita…

cs.LG2024

Data Efficient Evaluation of Large Language Models and Text-to-Image Models via Adaptive Sampling

Cong Xu, Gayathri Saranathan, Mahammad Parwez Alam +5

Evaluating LLMs and text-to-image models is a computationally intensive task often overlooked. Efficient evaluation is crucial for understanding the diverse capabilities of these m…

cs.CV2026

AIM 2025 Rip Current Segmentation (RipSeg) Challenge Report

Andrei Dumitriu, Florin Miron, Florin Tatui +24

This report presents an overview of the AIM 2025 RipSeg Challenge, a competition designed to advance techniques for automatic rip current segmentation in still images. Rip currents…

astro-ph2001

Models for Multiband IR Surveys

Cong Xu, Carol J. Lonsdale, David L. Shupe +2

Empirical 'backward' galaxy evolution models for IR-bright galaxies are constrained using multiband IR surveys. A new Monte-Carlo algorithm is developed for this task. It exploits…

astro-ph1998

Emission Features and Source Counts of Galaxies in Mid-Infrared

Cong Xu, Perry B. Hacking, Fan Fang +6

In this work we incorporate the newest ISO results on the mid-infrared spectral-energy-distributions (MIR SEDs) of galaxies into models for the number counts and redshift distribut…

cs.CV2023

Inner-IoU: More Effective Intersection over Union Loss with Auxiliary Bounding Box

Hao Zhang, Cong Xu, Shuaijie Zhang

With the rapid development of detectors, Bounding Box Regression (BBR) loss function has constantly updated and optimized. However, the existing IoU-based BBR still focus on accele…

physics.soc-ph2025

Spreading dynamics of information on online social networks

Fanhui Meng, Jiarong Xie, Jiachen Sun +7

Social media is profoundly changing our society with its unprecedented spreading power. Due to the complexity of human behaviors and the diversity of massive messages, the informat…

cs.CV2022

Adversarial Momentum-Contrastive Pre-Training

Cong Xu, Dan Li, Min Yang

Recently proposed adversarial self-supervised learning methods usually require big batches and long training epochs to extract robust features, which will bring heavy computational…

cs.IR2024

STAIR: Manipulating Collaborative and Multimodal Information for E-Commerce Recommendation

Cong Xu, Yunhang He, Jun Wang +1

While the mining of modalities is the focus of most multimodal recommendation methods, we believe that how to fully utilize both collaborative and multimodal information is pivotal…

cs.CV2024

Infer Induced Sentiment of Comment Response to Video: A New Task, Dataset and Baseline

Qi Jia, Baoyu Fan, Cong Xu +7

Existing video multi-modal sentiment analysis mainly focuses on the sentiment expression of people within the video, yet often neglects the induced sentiment of viewers while watch…

cs.CV2021

An Orthogonal Classifier for Improving the Adversarial Robustness of Neural Networks

Cong Xu, Xiang Li, Min Yang

Neural networks are susceptible to artificially designed adversarial perturbations. Recent efforts have shown that imposing certain modifications on classification layer can improv…

astro-ph2000

Mapping IR Enhancements in Closely Interacting Spiral-Spiral Pairs. I. ISO~CAM and ISO~SWS Observations

Cong Xu, Yu Gao, Joseph Mazzarella +3

Mid-infrared (MIR) imaging and spectroscopic observations are presented for a well defined sample of eight closely interacting (CLO) pairs of spiral galaxies that have overlapping…

astro-ph1995

High Resolution IRAS Maps and IR Emission of M31 --- I. Morphology and Sources

Cong Xu, George Helou

The morphology of the IR emission and the properties of discrete FIR sources in M31 disk are studied using the HiRes maps from IRAS. Very thin and bright FIR arm segments are shown…

cs.IR2025

Pattern-wise Transparent Sequential Recommendation

Kun Ma, Cong Xu, Zeyuan Chen +1

A transparent decision-making process is essential for developing reliable and trustworthy recommender systems. For sequential recommendation, it means that the model can identify…

cs.LG2025

Pushing the Limits of Low-Bit Optimizers: A Focus on EMA Dynamics

Cong Xu, Wenbin Liang, Mo Yu +7

The rapid scaling of models has led to prohibitively high training and fine-tuning costs. A major factor accounting for memory consumption is the widespread use of stateful optimiz…

cs.CV2024

Comparing remote sensing-based forest biomass mapping approaches using new forest inventory plots in contrasting forests in northeastern and southwestern China

Wenquan Dong, Edward T. A. Mitchard, Yuwei Chen +5

Large-scale high spatial resolution aboveground biomass (AGB) maps play a crucial role in determining forest carbon stocks and how they are changing, which is instrumental in under…

cs.LG2025

CAdam: Confidence-Based Optimization for Online Learning

Shaowen Wang, Anan Liu, Jian Xiao +9

Modern recommendation systems frequently employ online learning to dynamically update their models with freshly collected data. The most commonly used optimizer for updating neural…

quant-ph2026

Quantum average correlations and complementarity relations via metric-adjusted skew information

Xiaoyu Ma, Qing-Hua Zhang, Cong Xu

We investigate quantum average correlations and complementarity relations based on metric-adjusted skew information. Several natural averaging procedures are considered, including…

physics.soc-ph2024

Reconstructing the evolution history of networked complex systems

Junya Wang, Yi-Jiao Zhang, Cong Xu +6

The evolution processes of complex systems carry key information in the systems' functional properties. Applying machine learning algorithms, we demonstrate that the historical for…

cs.IR2024

Understanding the Role of Cross-Entropy Loss in Fairly Evaluating Large Language Model-based Recommendation

Cong Xu, Zhangchi Zhu, Jun Wang +2

Large language models (LLMs) have gained much attention in the recommendation community; some studies have observed that LLMs, fine-tuned by the cross-entropy loss with a full soft…

stat.AP2023

Where to serve and return in Badminton Men's Double?

Xuelin Zhu, Yu Sun, Yumin Zeng +1

This study aims to analyze the service and return landing areas in badminton men's double, based on data extracted from 20 badminton matches. We find that most services land near t…

cs.CL2026

Beyond Semantic Understanding: Preserving Collaborative Frequency Components in LLM-based Recommendation

Minhao Wang, Yunhang He, Cong Xu +4

Recommender systems in concert with Large Language Models (LLMs) present promising avenues for generating semantically-informed recommendations. However, LLM-based recommenders exh…

quant-ph2022

Uncertainty of quantum channels via modified generalized variance and modified generalized Wigner-Yanase-Dyson skew information

Cong Xu, Zhaoqi Wu, Shao-Ming Fei

Uncertainty relation is a fundamental issue in quantum mechanics and quantum information theory. By using modified generalized variance (MGV), and modified generalized Wigner-Yanas…

astro-ph2000

Molecular Gas Concentrations Outside the Merging Disks

Yu Gao, J. D. Goldader, E. R. Seaquist +1

We present our preliminary BIMA CO(1-0) images of II~Zw~96 which show huge molecular gas concentrations outside the merging disks. The dominant extra-disk CO concentrations in II~Z…

cs.IR2026

ASPIRE: Make Spectral Graph Collaborative Filtering Great Again via Adaptive Filter Learning

Yunhang He, Cong Xu, Zhangchi Zhu +2

Graph filter design is central to spectral collaborative filtering, yet most existing methods rely on manually tuned hyperparameters rather than fully learnable filters. We show th…

astro-ph2003

SWIRE: The SIRTF Wide-area InfraRed Extragalactic Survey

Carol J. Lonsdale, Harding E. Smith, Michael Rowan-Robinson +16

The SIRTF Wide-area InfraRed Extragalactic survey (SWIRE), the largest SIRTF Legacy program, is a wide-area, imaging survey to trace the evolution of dusty, star-forming galaxies,…

cs.IR2026

HiGR: Industrial-Scale Hierarchical Generative Slate Recommendation Framework in Tencent

Yunsheng Pang, Zijian Liu, Yudong Li +10

Slate recommendation, which presents users with a ranked item list in a single display, is ubiquitous across mainstream online platforms. While recent generative recommendation met…

cs.CV2025

DropletVideo: A Dataset and Approach to Explore Integral Spatio-Temporal Consistent Video Generation

Runze Zhang, Guoguang Du, Xiaochuan Li +10

Spatio-temporal consistency is a critical research topic in video generation. A qualified generated video segment must ensure plot plausibility and coherence while maintaining visu…

gr-qc2026

Testing the Transverse Scalar Mode of Gravitational Quantum Field Theory with Taiji and LISA

Cong Xu, Yong Tang, Yue-Liang Wu

The paper investigates how the transverse scalar (breathing) polarization predicted by Gravitational Quantum Field Theory can be detected with space‑based gravitational‑wave detect…

#gravitational waves#alternative theories of gravity#polarization modes#space-based detectors
stat.ML2019

A Fast deflation Method for Sparse Principal Component Analysis via Subspace Projections

Cong Xu, Min Yang, Jin Zhang

The implementation of conventional sparse principal component analysis (SPCA) on high-dimensional data sets has become a time consuming work. In this paper, a series of subspace pr…

astro-ph2000

Local Luminosity Function at 15 and Galaxy Evolution Seen by ISOCAM 15 Surveys

Cong Xu

A local luminosity function at 15 is derived using the bivariate (15 vs. 60 luminosity) method, based on the newly published ISOCAM LW3-band (15) survey of the…

cs.AI2026

Explainable Knowledge Tracing via Probabilistic Embeddings and Pattern-based Reasoning

Siyu Wu, Cong Xu, Wei Zhang

Knowledge Tracing (KT) models students' knowledge states based on learning interactions to predict performance. While deep learning-based KT models have boosted predictive accuracy…

astro-ph2000

CO in Stephan's Quintet: First Evidence of Molecular Gas in the Intragroup Starburst

Yu Gao, Cong Xu

We present the first interferometric evidence of CO(1-0) emission in the intragroup starburst (IGS) region of HCG~92 (Stephan's Quintet). BIMA's large primary beam covers fully bot…

cs.CV2025

MARS2 2025 Challenge on Multimodal Reasoning: Datasets, Methods, Results, Discussion, and Outlook

Peng Xu, Shengwu Xiong, Jiajun Zhang +125

This paper reviews the MARS2 2025 Challenge on Multimodal Reasoning. We aim to bring together different approaches in multimodal machine learning and LLMs via a large benchmark. We…

quant-ph2022

Sum uncertainty relations based on weighted Wigner-Yanase-Dyson skew information

Cong Xu, Zhaoqi Wu, Shao-Ming Fei

We introduce () weighted Wigner-Yanase-Dyson (() WWYD) skew information and () modified weighted Wigner-Yanase-Dyson (() MWWYD) skew informa…

cs.CV2017

Coordinating Filters for Faster Deep Neural Networks

Wei Wen, Cong Xu, Chunpeng Wu +3

Very large-scale Deep Neural Networks (DNNs) have achieved remarkable successes in a large variety of computer vision tasks. However, the high computation intensity of DNNs makes i…

eess.SP2024

f-GAN: A frequency-domain-constrained generative adversarial network for PPG to ECG synthesis

Nathan C. L. Kong, Dae Lee, Huyen Do +4

Electrocardiograms (ECGs) and photoplethysmograms (PPGs) are generally used to monitor an individual's cardiovascular health. In clinical settings, ECGs and fingertip PPGs are the…

cs.IR2024

Graph-enhanced Optimizers for Structure-aware Recommendation Embedding Evolution

Cong Xu, Jun Wang, Jianyong Wang +1

Embedding plays a key role in modern recommender systems because they are virtual representations of real-world entities and the foundation for subsequent decision-making models. I…

cs.IR2025

EnhancedRL: An Enhanced-State Reinforcement Learning Algorithm for Multi-Task Fusion in Recommender Systems

Peng Liu, Cong Xu, Jiawei Zhu +2

As a key stage of Recommender Systems (RSs), Multi-Task Fusion (MTF) is responsible for merging multiple scores output by Multi-Task Learning (MTL) into a single score, finally det…

astro-ph1998

The Mid-Infrared Color-Luminosity Relation and the Local 12 micron Luminosity Function

Fan Fang, David L. Shupe, Cong Xu +1

We have established a model to systematically estimate the contribution of the mid-infrared emission features between 3 and 11.6 micron to the IRAS in-band fluxes, using the result…

quant-ph2025

Quantifying nonclassical correlations relative to local channels

Cong Xu, Tao Li, Ruonan Ren +2

Nonclassical correlations are significant physical resources with extensive applications in quantum information processing. We introduce the modified Wigner-Yanase-Dyson skew infor…

cs.CV2024

Understanding Adversarial Robustness from Feature Maps of Convolutional Layers

Cong Xu, Wei Zhang, Jun Wang +1

The adversarial robustness of a neural network mainly relies on two factors: model capacity and anti-perturbation ability. In this paper, we study the anti-perturbation ability of…

cs.LG2023

Less Emphasis on Difficult Layer Regions: Curriculum Learning for Singularly Perturbed Convection-Diffusion-Reaction Problems

Yufeng Wang, Cong Xu, Min Yang +1

Although Physics-Informed Neural Networks (PINNs) have been successfully applied in a wide variety of science and engineering fields, they can fail to accurately predict the underl…

quant-ph2026

Quantum average correlation based on average coherence

Xiaoyu Ma, Qing-Hua Zhang, Cong Xu

This paper studies the quantification and structural properties of quantum average correlation based on average coherence. Motivated by two mathematically equivalent approaches to…

quant-ph2023

Tighter sum uncertainty relations via weighted Wigner-Yanase-Dyson skew information

Cong Xu, Zhaoqi Wu, Shao-Ming Fei

We establish tighter uncertainty relations for arbitrary finite observables via weighted Wigner-Yanase-Dyson (WWYD) skew information. The results are also…

physics.soc-ph2025

Critical Thresholds in Non-Pharmaceutical Interventions for Epidemic Control

Jinghui Wang, Yutian Zeng, Cong Xu +7

Non-pharmaceutical interventions, such as contact tracing and social distancing, are critical for controlling epidemic outbreaks, yet their dynamic interactions remain underexplore…

quant-ph2024

Uncertainty relations based on the -absolute variance for quantum channels

Cong Xu, Wen Zhou, Qing-Hua Zhang +1

Uncertainty principle reveals the intrinsic differences between the classical and quantum worlds, which plays a significant role in quantum information theory. By using -absolu…

cs.CV2026

Bridging the Generalization Gap in Adverse Weather Segmentation: A Training Recipe Perspective

Cong Xu, Pu Luo, Yumei Li +1

This paper describes our approach for the 8th UG2+ Workshop (CVPR 2026) Track~2, which targets semantic segmentation of outdoor scenes degraded by five weather conditions: blur, da…

quant-ph2025

Tightening the entropic uncertainty relations with quantum memory in a multipartite scenario

Cong Xu, Qing-Hua Zhang, Tao Li +1

The quantum uncertainty principle stands as a cornerstone and a distinctive feature of quantum mechanics, setting it apart from classical mechanics. We introduce a tripartite quant…

quant-ph2023

Summation and product forms of uncertainty relations based on metric-adjusted skew information

Cong Xu, Qing-Hua Zhang, Shao-Ming Fei

Uncertainty principle is one of the most essential features in quantum mechanics and plays profound roles in quantum information processing. We establish tighter summation form unc…

eess.IV2022

Sparse-based Domain Adaptation Network for OCTA Image Super-Resolution Reconstruction

Huaying Hao, Cong Xu, Dan Zhang +4

Retinal Optical Coherence Tomography Angiography (OCTA) with high-resolution is important for the quantification and analysis of retinal vasculature. However, the resolution of OCT…

quant-ph2026

Tighter entropic uncertainty relations in the presence of quantum memories for complete sets of mutually unbiased bases

Qing-Hua Zhang, Cong Xu, Jing-Feng Wu +1

Entropic uncertainty relations provide an information-theoretic framework for quantifying the fundamental indeterminacy inherent in quantum mechanics. We propose more stringent qua…

cs.IR2025

Dynamic User Interest Augmentation via Stream Clustering and Memory Networks in Large-Scale Recommender Systems

Peng Liu, Nian Wang, Cong Xu +3

Recommender System (RS) provides personalized recommendation service based on user interest. However, lots of users' interests are sparse due to lacking consumption behaviors, maki…

cs.IR2025

Collaborative Filtering Meets Spectrum Shift: Connecting User-Item Interaction with Graph-Structured Side Information

Yunhang He, Cong Xu, Jun Wang +1

Graph Neural Networks (GNNs) have demonstrated their superiority in collaborative filtering, where the user-item (U-I) interaction bipartite graph serves as the fundamental data fo…

cs.IR2025

UnifiedRL: A Reinforcement Learning Algorithm Tailored for Multi-Task Fusion in Large-Scale Recommender Systems

Peng Liu, Cong Xu, Ming Zhao +3

As the last pivotal stage of Recommender System (RS), Multi-Task Fusion (MTF) is responsible for combining multiple scores outputted by Multi-Task Learning (MTL) model into a final…

cs.IR2026

Markovian Pre-Trained Transformer for Next-Item Recommendation

Cong Xu, Guoliang Li, Jun Wang +1

We introduce the Markovian Pre-trained Transformer (MPT) for next-item recommendation, a transferable model fully pre-trained on synthetic Markov chains, yet capable of achieving s…