arXiv subjects

Yi Xu

Publications and source records attributed to Yi Xu.

At least 181 records · Page 10Linked to original sources

arXiv2025

SimMIL: A Universal Weakly Supervised Pre-Training Framework for Multi-Instance Learning in Whole Slide Pathology Images

Various multi-instance learning (MIL) based approaches have been developed and successfully applied to whole-slide pathological images (WSI). Existing MIL methods emphasize the importance of feature aggregators, but largely neglect the instance-level representation learning. They assume that the availability of a pre-trained feature extractor can be directly utilized or fine-tuned, which is not always the case. This paper proposes to pre-train feature extractor for MIL via a weakly-supervised scheme, i.e., propagating the weak bag-level labels to the corresponding instances for supervised learning. To learn effective features for MIL, we further delve into several key components, including strong data augmentation, a non-linear prediction head and the robust loss function. We conduct experiments on common large-scale WSI datasets and find it achieves better performance than other pre-training schemes (e.g., ImageNet pre-training and self-supervised learning) in different downstream tasks. We further show the compatibility and scalability of the proposed scheme by deploying it in fine-tuning the pathological-specific models and pre-training on merged multiple datasets. To our knowledge, this is the first work focusing on the representation learning for MIL.

cs.CV↗

arXiv2025

RGB-D Tracking via Hierarchical Modality Aggregation and Distribution Network

The integration of dual-modal features has been pivotal in advancing RGB-Depth (RGB-D) tracking. However, current trackers are less efficient and focus solely on single-level features, resulting in weaker robustness in fusion and slower speeds that fail to meet the demands of real-world applications. In this paper, we introduce a novel network, denoted as HMAD (Hierarchical Modality Aggregation and Distribution), which addresses these challenges. HMAD leverages the distinct feature representation strengths of RGB and depth modalities, giving prominence to a hierarchical approach for feature distribution and fusion, thereby enhancing the robustness of RGB-D tracking. Experimental results on various RGB-D datasets demonstrate that HMAD achieves state-of-the-art performance. Moreover, real-world experiments further validate HMAD's capacity to effectively handle a spectrum of tracking challenges in real-time scenarios.

cs.CV↗

arXiv2025

Efficient Data Valuation Approximation in Federated Learning: A Sampling-based Approach

Federated learning paradigm to utilize datasets across multiple data providers. In FL, cross-silo data providers often hesitate to share their high-quality dataset unless their data value can be fairly assessed. Shapley value (SV) has been advocated as the standard metric for data valuation in FL due to its desirable properties. However, the computational overhead of SV is prohibitive in practice, as it inherently requires training and evaluating an FL model across an exponential number of dataset combinations. Furthermore, existing solutions fail to achieve high accuracy and efficiency, making practical use of SV still out of reach, because they ignore choosing suitable computation scheme for approximation framework and overlook the property of utility function in FL. We first propose a unified stratified-sampling framework for two widely-used schemes. Then, we analyze and choose the more promising scheme under the FL linear regression assumption. After that, we identify a phenomenon termed key combinations, where only limited dataset combinations have a high-impact on final data value. Building on these insights, we propose a practical approximation algorithm, IPSS, which strategically selects high-impact dataset combinations rather than evaluating all possible combinations, thus substantially reducing time cost with minor approximation error. Furthermore, we conduct extensive evaluations on the FL benchmark datasets to demonstrate that our proposed algorithm outperforms a series of representative baselines in terms of efficiency and effectiveness.

cs.LG↗

arXiv2025

A Galaxy with an Extremely Blue UV Slope $β=-3$ at $z=9.25$ Identified by JWST Spectroscopy: Evidence for a Weak Nebular Continuum and Efficient Ionizing Photon Escape?

We investigate UV continuum slopes $β$ of 863 galaxies at $z=4-14$ using archival JWST/NIRSpec PRISM spectra obtained from major JWST GTO, ERS, and GO programs, including JADES, CEERS, and UNCOVER. Among these galaxies, we identify a remarkable galaxy at $z=9.25$, dubbed EBG-1, with a significantly blue UV slope $β=-2.99\pm0.15$, unlike the rest of the galaxies that exhibit red continua or ambiguous blue continua hindered by large uncertainties. We confirm that the $β$ value negligibly changes by the data reduction and fitting wavelength ranges for UV emission/absorption line masking. The extreme blue slope, $β=-3.0$, rules out significant contributions from dust extinction or AGN activity. Comparing with stellar and nebular emission models, we find that such a blue UV slope cannot be reproduced solely by stellar models even with very young, metal-poor, or top-heavy contiguous star formation associated with strong nebular continua making the UV slopes red, but with a high ionizing photon escape fraction, $f_\mathrm{esc}^\mathrm{ion} \gtrsim 0.5$, for a weak nebular continuum. While the H$β$ emission line is not detected, likely due to the limited sensitivity of the spectrum, we find moderately weak [O III] $λλ$4959,5007 emission lines for the given star-formation rate ($3\, \mathrm{M_\odot}$ yr$^{-1}$) and stellar mass ($10^{8.0} \, \mathrm{M_\odot}$) that are about three times weaker than the average emission lines, again suggestive of the high ionizing photon escape fraction, $f_\mathrm{esc}^\mathrm{ion} \sim 0.7$ or more. EBG-1 would provide crucial insights into stellar and nebular continuum emission in high-redshift galaxies, serving as an example of the ionizing photon escaping site at the epoch of reionization.

astro-ph.GA↗

arXiv2025

Quasi-Irreducibility of Nonnegative Biquadratic Tensors

While the adjacency tensor of a bipartite 2-graph is a nonnegative biquadratic tensor, it is inherently reducible. To address this limitation, we introduce the concept of quasi-irreducibility in this paper. The adjacency tensor of a bipartite 2-graph is quasi-irreducible if that bipartite 2-graph is not bi-separable. This new concept reveals important spectral properties: although all M$^+$-eigenvalues are M$^{++}$-eigenvalues for irreducible nonnegative biquadratic tensors, the M$^+$-eigenvalues of a quasi-irreducible nonnegative biquadratic tensor can be either M$^0$-eigenvalues or M$^{++}$-eigenvalues. Furthermore, we establish a max-min theorem for the M-spectral radius of a nonnegative biquadratic tensor.

math.SP↗

arXiv2025

PanoDreamer: Consistent Text to 360-Degree Scene Generation

Automatically generating a complete 3D scene from a text description, a reference image, or both has significant applications in fields like virtual reality and gaming. However, current methods often generate low-quality textures and inconsistent 3D structures. This is especially true when extrapolating significantly beyond the field of view of the reference image. To address these challenges, we propose PanoDreamer, a novel framework for consistent, 3D scene generation with flexible text and image control. Our approach employs a large language model and a warp-refine pipeline, first generating an initial set of images and then compositing them into a 360-degree panorama. This panorama is then lifted into 3D to form an initial point cloud. We then use several approaches to generate additional images, from different viewpoints, that are consistent with the initial point cloud and expand/refine the initial point cloud. Given the resulting set of images, we utilize 3D Gaussian Splatting to create the final 3D scene, which can then be rendered from different viewpoints. Experiments demonstrate the effectiveness of PanoDreamer in generating high-quality, geometrically consistent 3D scenes.

cs.CV↗

arXiv2025

Decoding Recommendation Behaviors of In-Context Learning LLMs Through Gradient Descent

Recently, there has been a growing trend in utilizing large language models (LLMs) for recommender systems, referred to as LLMRec. A notable approach within this trend is not to fine-tune these models directly but instead to leverage In-Context Learning (ICL) methods tailored for LLMRec, denoted as LLM-ICL Rec. Many contemporary techniques focus on harnessing ICL content to enhance LLMRec performance. However, optimizing LLMRec with ICL content presents unresolved challenges. Specifically, two key issues stand out: (1) the limited understanding of why using a few demonstrations without model fine-tuning can lead to better performance compared to zero-shot recommendations. (2) the lack of evaluation metrics for demonstrations in LLM-ICL Rec and the absence of the theoretical analysis and practical design for optimizing the generation of ICL content for recommendation contexts. To address these two main issues, we propose a theoretical model, the LLM-ICL Recommendation Equivalent Gradient Descent model (LRGD) in this paper, which connects recommendation generation with gradient descent dynamics. We demonstrate that the ICL inference process in LLM aligns with the training procedure of its dual model, producing token predictions equivalent to the dual model's testing outputs. Building on these theoretical insights, we propose an evaluation metric for assessing demonstration quality. We integrate perturbations and regularizations in LRGD to enhance the robustness of the recommender system. To further improve demonstration effectiveness, prevent performance collapse, and ensure long-term adaptability, we also propose a two-stage optimization process in practice. Extensive experiments and detailed analysis on three Amazon datasets validate the theoretical equivalence and support the effectiveness of our theoretical analysis and practical module design.

cs.IR↗

arXiv2025

Unified Convergence Analysis for Adaptive Optimization with Moving Average Estimator

Although adaptive optimization algorithms have been successful in many applications, there are still some mysteries in terms of convergence analysis that have not been unraveled. This paper provides a novel non-convex analysis of adaptive optimization to uncover some of these mysteries. Our contributions are three-fold. First, we show that an increasing or large enough momentum parameter for the first-order moment used in practice is sufficient to ensure the convergence of adaptive algorithms whose adaptive scaling factors of the step size are bounded. Second, our analysis gives insights for practical implementations, e.g., increasing the momentum parameter in a stage-wise manner in accordance with stagewise decreasing step size would help improve the convergence. Third, the modular nature of our analysis allows its extension to solving other optimization problems, e.g., compositional, min-max and bilevel problems. As an interesting yet non-trivial use case, we present algorithms for solving non-convex min-max optimization and bilevel optimization that do not require using large batches of data to estimate gradients or double loops as the literature do. Our empirical studies corroborate our theoretical results.

math.OC↗

arXiv2025

IBCB: Efficient Inverse Batched Contextual Bandit for Behavioral Evolution History

Traditional imitation learning focuses on modeling the behavioral mechanisms of experts, which requires a large amount of interaction history generated by some fixed expert. However, in many streaming applications, such as streaming recommender systems, online decision-makers typically engage in online learning during the decision-making process, meaning that the interaction history generated by online decision-makers includes their behavioral evolution from novice expert to experienced expert. This poses a new challenge for existing imitation learning approaches that can only utilize data from experienced experts. To address this issue, this paper proposes an inverse batched contextual bandit (IBCB) framework that can efficiently perform estimations of environment reward parameters and learned policy based on the expert's behavioral evolution history. Specifically, IBCB formulates the inverse problem into a simple quadratic programming problem by utilizing the behavioral evolution history of the batched contextual bandit with inaccessible rewards. We demonstrate that IBCB is a unified framework for both deterministic and randomized bandit policies. The experimental results indicate that IBCB outperforms several existing imitation learning algorithms on synthetic and real-world data and significantly reduces running time. Additionally, empirical analyses reveal that IBCB exhibits better out-of-distribution generalization and is highly effective in learning the bandit policy from the interaction history of novice experts.

cs.LG↗

arXiv2025

Stellar- and AGN-Driven Outflows in JWST Galaxies at z=3-9: More Frequent, Wider Opening Angles, and Mostly Bounded

We study outflows in 130 galaxies with -22 1000 km s$^{-1}$ while one galaxy is identified as a Type-2 AGN by high ionization emission lines. With the velocity shift and line widths of the outflow broad lines, we obtain ~80-500 km s$^{-1}$ for the outflow velocities. We find that the outflow velocities are slower than low-z galaxies with similar SFRs, which may be explained by the low stellar masses of high-z galaxies. The outflow velocities of AGNs are large but not significantly different from the others. Interestingly, these outflow velocities are typically not high enough to escape from the galactic potentials, possibly suggesting fountain-type outflows. We estimate mass loading factors $η$ to be 0.1-1 that are not particularly large, but comparable with those of z~1 outflows. The large fraction of galaxies with outflows (30% with high resolution data) provides constraints on outflow parameters, suggesting a wide opening angle of >45 deg and a large duty-cycle of >30%, which gives a picture of more frequent and spherical outflows in high-z galaxies.

astro-ph.GA↗

arXiv2025

Diagnosis of Pulmonary Hypertension by Integrating Multimodal Data with a Hybrid Graph Convolutional and Transformer Network

Early and accurate diagnosis of pulmonary hypertension (PH) is essential for optimal patient management. Differentiating between pre-capillary and post-capillary PH is critical for guiding treatment decisions. This study develops and validates a deep learning-based diagnostic model for PH, designed to classify patients as non-PH, pre-capillary PH, or post-capillary PH. This retrospective study analyzed data from 204 patients (112 with pre-capillary PH, 32 with post-capillary PH, and 60 non-PH controls) at the First Affiliated Hospital of Nanjing Medical University. Diagnoses were confirmed through right heart catheterization. We selected 6 samples from each category for the test set (18 samples, 10%), with the remaining 186 samples used for the training set. This process was repeated 35 times for testing. This paper proposes a deep learning model that combines Graph convolutional networks (GCN), Convolutional neural networks (CNN), and Transformers. The model was developed to process multimodal data, including short-axis (SAX) sequences, four-chamber (4CH) sequences, and clinical parameters. Our model achieved a performance of Area under the receiver operating characteristic curve (AUC) = 0.81 +- 0.06(standard deviation) and Accuracy (ACC) = 0.73 +- 0.06 on the test set. The discriminative abilities were as follows: non-PH subjects (AUC = 0.74 +- 0.11), pre-capillary PH (AUC = 0.86 +- 0.06), and post-capillary PH (AUC = 0.83 +- 0.10). It has the potential to support clinical decision-making by effectively integrating multimodal data to assist physicians in making accurate and timely diagnoses.

eess.IV↗

arXiv2025

Slitless Areal Pure-Parallel HIgh-Redshift Emission Survey (SAPPHIRES): Early Data Release of Deep JWST/NIRCam Images and Spectra in MACS J0416 Parallel Field

We present the early data release (EDR) of SAPPHIRES, a JWST Cycle-3 Treasury imaging and spectroscopic survey using the powerful NIRCam wide-field slitless spectroscopic (WFSS) mode in pure parallel. SAPPHIRES will obtain NIRCam imaging and WFSS data in many cosmological deep fields totaling a telescope charged time of 709 hours (557-hour exposures). In this EDR, we present NIRCam imaging and WFSS data obtained in parallel to the Frontier Field galaxy cluster MACS J0416.1-2403, which are attached to primary observations JWST-GO-4750. With a total dual-channel exposure time of 47.2 hours, we obtain deep NIRCam imaging in 13 bands at 0.6--5.0 micron and deep WFSS at 3.1-5.0 micron through the F356W and F444W filters with grisms in orthogonal dispersion directions. We release reduced NIRCam images, photometric catalogs of 22107 sources and WFSS spectra of 1060 sources with confirmed redshifts ($z\simeq0-8.5$). Preliminary value-added catalogs including photometric redshifts, spectroscopic redshifts and physical properties (mass, star-formation rate, etc.) are also made available. We also characterize the data quality and demonstrate scientific applications, including (1) galaxy candidates at the redshift frontier ($z\gtrsim10$), (2) the ionized gas kinematics of a galaxy reconstructed from $R\sim1500$ grism spectra at orthogonal dispersion directions, (3) massive emission-line galaxies and active galactic nuclei (AGN) around the Epoch of Reionization.

astro-ph.GA↗

arXiv2025

SAPPHIRES: A Galaxy Over-Density in the Heart of Cosmic Reionization at $z=8.47$

We report the discovery of a galaxy proto-cluster candidate (dubbed MACS0416-OD-z8p5) at a spectroscopic redshift of $z\sim8.47$, dating back to $\sim550$Myr after the Big Bang. The observations are part of the JWST Cycle-3 treasury program, Slitless Areal Pure-Parallel HIgh-Redshift Emission Survey (SAPPHIRES) with NIRCam-grism. Using wide field slitless spectroscopy (WFSS) obtained in the MACS0416 parallel field, we robustly confirm nine galaxies at $z_{\rm spec}\sim8.47$ via emission line detections of [OIII]5008A (with $>5\,σ$) and tentatively confirm one additional galaxy (at $\sim3\,σ$). This discovery represents the highest-redshift, spectroscopically confirmed galaxy over-density known to date, which is $\sim6$--$8$ times more dense than the average volume density of galaxies at the same redshift. Furthermore, a galaxy hosting a low-mass active galactic nucleus (``Little-Red-Dot'') is found as a member, suggesting an early emergence of active, massive black holes and feedback between these black holes and their surrounding environments. We also discuss the spatial structures connecting the galaxy over-density to nearby massive star-forming galaxies (separated by $\sim 5$pMpc, including MACS0416-Y1 and MACS0416-JD. This finding of a massive dark matter halo hosting a galaxy over-density at $z\sim8.5$ is surprising given that our survey covered only a small, random field ($16.5\,{\rm arcmin^2}$) as part of a pure parallel observation. The comparison with cosmological simulations shows that the likelihood of finding such a large-scale structure is $<5\,\%$ under the current galaxy formation scenario and the observed survey volume. Our results demonstrate the power of WFSS observations to build a complete line-emitter sample and suggest an important role for over-densities in enhancing galaxy formation by funneling large-scale gas supplies into small cosmological volumes.

astro-ph.GA↗

arXiv2025

Addressing Information Loss and Interaction Collapse: A Dual Enhanced Attention Framework for Feature Interaction

The Transformer has proven to be a significant approach in feature interaction for CTR prediction, achieving considerable success in previous works. However, it also presents potential challenges in handling feature interactions. Firstly, Transformers may encounter information loss when capturing feature interactions. By relying on inner products to represent pairwise relationships, they compress raw interaction information, which can result in a degradation of fidelity. Secondly, due to the long-tail features distribution, feature fields with low information-abundance embeddings constrain the information abundance of other fields, leading to collapsed embedding matrices. To tackle these issues, we propose a Dual Attention Framework for Enhanced Feature Interaction, known as Dual Enhanced Attention. This framework integrates two attention mechanisms: the Combo-ID attention mechanism and the collapse-avoiding attention mechanism. The Combo-ID attention mechanism directly retains feature interaction pairs to mitigate information loss, while the collapse-avoiding attention mechanism adaptively filters out low information-abundance interaction pairs to prevent interaction collapse. Extensive experiments conducted on industrial datasets have shown the effectiveness of Dual Enhanced Attention.

cs.IR↗

arXiv2025

The Physical Origin of Extreme Emission Line Galaxies at High redshifts: Strong {\sc [Oiii]} Emission Lines Produced by Obscured AGNs

We present deep Subaru/FOCAS spectra for two extreme emission line galaxies (EELGs) at $z\sim 1$ with strong {\sc[Oiii]}$λ$5007 emission lines, exhibiting equivalent widths (EWs) of $2905^{+946}_{-578}$ Å and $2000^{+188}_{-159}$ Å, comparable to those of EELGs at high redshifts that are now routinely identified with JWST spectroscopy. Adding a similarly large {\sc [Oiii]} EW ($2508^{+1487}_{-689}$ Å) EELG found at $z\sim 2$ in the JWST CEERS survey to our sample, we explore for the physical origins of the large {\sc [Oiii]} EWs of these three galaxies with the Subaru spectra and various public data including JWST/NIRSpec, NIRCam, and MIRI data. While there are no clear signatures of AGN identified by the optical line diagnostics, we find that two out of two galaxies covered by the MIRI data show strong near-infrared excess in the spectral energy distributions (SEDs) indicating obscured AGN. Because none of the three galaxies show clear broad H$β$ lines, the upper limits on the flux ratios of broad-H$β$ to {\sc [Oiii]} lines are small, $\lesssim 0.15$ that are comparable with Seyfert $1.8-2.0$ galaxies. We conduct \texttt{Cloudy} modeling with the stellar and AGN incident spectra, allowing a wide range of parameters including metallicities and ionization parameters. We find that the large {\sc [Oiii]} EWs are not self-consistently reproduced by the spectra of stars or unobscured AGN, but obscured AGN that efficiently produces O$^{++}$ ionizing photons with weak nuclear and stellar continua that are consistent with the SED shapes.

astro-ph.GA↗

arXiv2025

PlanarNeRF: Online Learning of Planar Primitives with Neural Radiance Fields

Identifying spatially complete planar primitives from visual data is a crucial task in computer vision. Prior methods are largely restricted to either 2D segment recovery or simplifying 3D structures, even with extensive plane annotations. We present PlanarNeRF, a novel framework capable of detecting dense 3D planes through online learning. Drawing upon the neural field representation, PlanarNeRF brings three major contributions. First, it enhances 3D plane detection with concurrent appearance and geometry knowledge. Second, a lightweight plane fitting module is proposed to estimate plane parameters. Third, a novel global memory bank structure with an update mechanism is introduced, ensuring consistent cross-frame correspondence. The flexible architecture of PlanarNeRF allows it to function in both 2D-supervised and self-supervised solutions, in each of which it can effectively learn from sparse training signals, significantly improving training efficiency. Through extensive experiments, we demonstrate the effectiveness of PlanarNeRF in various scenarios and remarkable improvement over existing works.

cs.CV↗

arXiv2025

EvolvingGS: High-Fidelity Streamable Volumetric Video via Evolving 3D Gaussian Representation

We have recently seen great progress in 3D scene reconstruction through explicit point-based 3D Gaussian Splatting (3DGS), notable for its high quality and fast rendering speed. However, reconstructing dynamic scenes such as complex human performances with long durations remains challenging. Prior efforts fall short of modeling a long-term sequence with drastic motions, frequent topology changes or interactions with props, and resort to segmenting the whole sequence into groups of frames that are processed independently, which undermines temporal stability and thereby leads to an unpleasant viewing experience and inefficient storage footprint. In view of this, we introduce EvolvingGS, a two-stage strategy that first deforms the Gaussian model to coarsely align with the target frame, and then refines it with minimal point addition/subtraction, particularly in fast-changing areas. Owing to the flexibility of the incrementally evolving representation, our method outperforms existing approaches in terms of both per-frame and temporal quality metrics while maintaining fast rendering through its purely explicit representation. Moreover, by exploiting temporal coherence between successive frames, we propose a simple yet effective compression algorithm that achieves over 50x compression rate. Extensive experiments on both public benchmarks and challenging custom datasets demonstrate that our method significantly advances the state-of-the-art in dynamic scene reconstruction, particularly for extended sequences with complex human performances.

cs.CV↗

arXiv2025

Sports-Traj: A Unified Trajectory Generation Model for Multi-Agent Movement in Sports

Understanding multi-agent movement is critical across various fields. The conventional approaches typically focus on separate tasks such as trajectory prediction, imputation, or spatial-temporal recovery. Considering the unique formulation and constraint of each task, most existing methods are tailored for only one, limiting the ability to handle multiple tasks simultaneously, which is a common requirement in real-world scenarios. Another limitation is that widely used public datasets mainly focus on pedestrian movements with casual, loosely connected patterns, where interactions between individuals are not always present, especially at a long distance, making them less representative of more structured environments. To overcome these limitations, we propose a Unified Trajectory Generation model, UniTraj, that processes arbitrary trajectories as masked inputs, adaptable to diverse scenarios in the domain of sports games. Specifically, we introduce a Ghost Spatial Masking (GSM) module, embedded within a Transformer encoder, for spatial feature extraction. We further extend recent State Space Models (SSMs), known as the Mamba model, into a Bidirectional Temporal Mamba (BTM) to better capture temporal dependencies. Additionally, we incorporate a Bidirectional Temporal Scaled (BTS) module to thoroughly scan trajectories while preserving temporal missing relationships. Furthermore, we curate and benchmark three practical sports datasets, Basketball-U, Football-U, and Soccer-U, for evaluation. Extensive experiments demonstrate the superior performance of our model. We hope that our work can advance the understanding of human movement in real-world applications, particularly in sports. Our datasets, code, and model weights are available here https://github.com/colorfulfuture/UniTraj-pytorch.

cs.CV↗