arXiv subjects

Hao Sun

Publications and source records attributed to Hao Sun.

At least 217 records · Page 12Linked to original sources

arXiv2025

Effects of inner crusts on $g$-mode oscillations in neutron stars

In this work we investigate the influence of neutron stars' crusts on the non-radial $g$-mode oscillations and examine their correlations with nuclear matter properties fixed by adopting 10 different relativistic density functionals. At subsaturation densities, neutron star matter takes non-uniform structures and form the crusts. We find that the Brunt-Väisälä (BV) frequency increases significantly at densities slightly above the neutron drip density (i.e., neutron stars' inner crusts), which leads to crust $g$-mode oscillations with their frequencies insensitive to the adopted density functional. At larger densities, BV frequency increases as well due to the core-crust transitions and emergence of muons, which lead to core $g$-mode oscillations. It is found that the obtained core $g$-mode frequencies generally increase with the slope of nuclear symmetry energy $L$, which eventually intersect with that of the crust $g$ modes adopting large enough $L$. This leads to the avoid-crossing phenomenon for the global $g$ modes that encompass contributions from both the crust and core. The correlation between the global $g_1$ mode and $L$ is identified for neutron stars with masses $M\gtrsim 1.4\ M_{\odot}$, which enables the measurements of $L$ based on gravitational wave observations. In our future study, the effects of the discontinuities in density or shear modulus should be considered, while the temperature, rotation, magnetic field, and superfluid neutron gas in neutron stars could also play important roles.

astro-ph.HE↗

arXiv2025

Demonstration of Direct-amplification Enabled Harmonic Generation in an Ultraviolet Free-Electron Laser

We report the experimental demonstration of direct-amplification enabled harmonic generation in an ultraviolet free-electron laser (FEL) driven by a low-intensity seed laser. By employing a versatile undulator configuration that enables seed amplification and harmonic generation within a unified setup, we achieved over 100-fold energy gain of the seed and observed exponential growth at the second harmonic. The results demonstrate that a sufficiently long modulator can not only amplify a weak seed but also induce strong energy modulation of the electron beam, enabling efficient harmonic bunching. This method markedly relaxes the power requirements on external seed lasers and presents a viable route toward high-repetition-rate, fully coherent FELs

physics.acc-ph↗

arXiv2025

Uncertainty-Guided Self-Questioning and Answering for Video-Language Alignment

The development of multi-modal models has been rapidly advancing, with some demonstrating remarkable capabilities. However, annotating video-text pairs remains expensive and insufficient. Take video question answering (VideoQA) tasks as an example, human annotated questions and answers often cover only part of the video, since the corresponding text is often short and monotonous, leading to underutilization of video. To address this, we propose a Bootstrapping Video-Language Alignment framework (BoViLA), a self-training method that augments question samples during training process through LLM-based self-questioning and answering, which help model exploit video information and the internal knowledge of LLMs more thoroughly to improve modality alignment. However, low-quality self-generated questions may instead contaminate the performance, especially in the early stages of training, as we have observed in our experiments. To filter bad self-generated questions, we introduce Evidential Deep Learning (EDL) to estimate uncertainty and assess the quality of self-generated questions by evaluating the modality alignment within the context. To the best of our knowledge, this work is the first to explore LLM-based self-training frameworks for modality alignment. We evaluate BoViLA on five strong VideoQA benchmarks, where it outperforms several state-of-the-art methods and demonstrate its effectiveness and generality. Additionally, we provide extensive analyses of the self-training framework and the EDL-based uncertainty filtering mechanism. The code will be made available.

cs.CV↗

arXiv2025

PeSANet: Physics-encoded Spectral Attention Network for Simulating PDE-Governed Complex Systems

Accurately modeling and forecasting complex systems governed by partial differential equations (PDEs) is crucial in various scientific and engineering domains. However, traditional numerical methods struggle in real-world scenarios due to incomplete or unknown physical laws. Meanwhile, machine learning approaches often fail to generalize effectively when faced with scarce observational data and the challenge of capturing local and global features. To this end, we propose the Physics-encoded Spectral Attention Network (PeSANet), which integrates local and global information to forecast complex systems with limited data and incomplete physical priors. The model consists of two key components: a physics-encoded block that uses hard constraints to approximate local differential operators from limited data, and a spectral-enhanced block that captures long-range global dependencies in the frequency domain. Specifically, we introduce a novel spectral attention mechanism to model inter-spectrum relationships and learn long-range spatial features. Experimental results demonstrate that PeSANet outperforms existing methods across all metrics, particularly in long-term forecasting accuracy, providing a promising solution for simulating complex systems with limited data and incomplete physics.

cs.LG↗

arXiv2025

Muyan-TTS: A Trainable Text-to-Speech Model Optimized for Podcast Scenarios with a $50K Budget

Recent advancements in text-to-speech (TTS) models have been driven by the integration of large language models (LLMs), enhancing semantic comprehension and improving speech naturalness. However, existing LLM-based TTS models often lack open-source training code and efficient inference acceleration frameworks, limiting their accessibility and adaptability. Additionally, there is no publicly available TTS model specifically optimized for podcast scenarios, which are in high demand for voice interaction applications. To address these limitations, we introduce Muyan-TTS, an open-source trainable TTS model designed for podcast applications within a $50,000 budget. Our model is pre-trained on over 100,000 hours of podcast audio data, enabling zero-shot TTS synthesis with high-quality voice generation. Furthermore, Muyan-TTS supports speaker adaptation with dozens of minutes of target speech, making it highly customizable for individual voices. In addition to open-sourcing the model, we provide a comprehensive data collection and processing pipeline, a full training procedure, and an optimized inference framework that accelerates LLM-based TTS synthesis. Our code and models are available at https://github.com/MYZY-AI/Muyan-TTS.

cs.SD↗

arXiv2025

Anti-Aesthetics: Protecting Facial Privacy against Customized Text-to-Image Synthesis

The rise of customized diffusion models has spurred a boom in personalized visual content creation, but also poses risks of malicious misuse, severely threatening personal privacy and copyright protection. Some studies show that the aesthetic properties of images are highly positively correlated with human perception of image quality. Inspired by this, we approach the problem from a novel and intriguing aesthetic perspective to degrade the generation quality of maliciously customized models, thereby achieving better protection of facial identity. Specifically, we propose a Hierarchical Anti-Aesthetic (HAA) framework to fully explore aesthetic cues, which consists of two key branches: 1) Global Anti-Aesthetics: By establishing a global anti-aesthetic reward mechanism and a global anti-aesthetic loss, it can degrade the overall aesthetics of the generated content; 2) Local Anti-Aesthetics: A local anti-aesthetic reward mechanism and a local anti-aesthetic loss are designed to guide adversarial perturbations to disrupt local facial identity. By seamlessly integrating both branches, our HAA effectively achieves the goal of anti-aesthetics from a global to a local level during customized generation. Extensive experiments show that HAA outperforms existing SOTA methods largely in identity removal, providing a powerful tool for protecting facial privacy and copyright.

cs.CV↗

arXiv2025

A two-component dark matter model with $Z_2 \times Z_4$ symmetry

We consider a two-component dark matter model with $Z_2 \times Z_4$ symmetry, where a singlet scalar $S$ and a Majorana fermion $χ$ are introduced as dark matter candidates. We also introduce another singlet scalar $S_0$ with a non-zero vacuum expectation value to the SM so that the fermion dark matter can obtain mass after spontaneous symmetry breaking. We have a new Higgs boson in the model and in the case of the decoupling limit, the fermion dark matter production is only determined by $S$ and the new Higgs boson. The mass hierarchy of these new particles can make a difference in the reaction rate of dark matter annihilation processes, contributing to different viable parameter spaces for different mass orderings. We randomly scanned the parameter space with six various cases under relic density constraint and found that when $χ$ is the lightest among the dark sector, $χ$ production is generated via the so-called forbidden channels. Moreover, we consider the combined limits arising from Higgs invisible decay, dark matter relic density and direct detection constraints. Within the chosen parameter space, direct detection results put the most stringent constraint, and we have a more flexible value for the scalar dark matter mass when the mass of $χ$ is not smaller than the new Higgs boson mass.

hep-ph↗

arXiv2025

Extremely strong spin-orbit coupling effect in light-element altermagnetic materials

Spin-orbit coupling is a key to realize many novel physical effects in condensed matter physics. Altermagnetic materials possess the duality of real-space antiferromagnetism and reciprocal-space ferromagnetism. It has not been explored that achieving strong spin-orbit coupling effect in light-element altermagnetic materials. In this work, based on symmetry analysis, the first-principles electronic structure calculations plus Dynamical Mean Field Theory, we demonstrate that there is strong spin-orbit coupling effect in light-element altermagnetic materials $\rm NiF_3$ and $\rm FeCO_3$, and then propose a mechanism for realizing such effective spin-orbit coupling. This mechanism reveals the cooperative effect of crystal symmetry, electron occupation, electronegativity, electron correlation, and intrinsic spin-orbit coupling. Our work provides an approach for searching light-element altermagnetic materials with an effective strong spin-orbit coupling.

cond-mat.mtrl-sci↗

arXiv2025

LL-Gaussian: Low-Light Scene Reconstruction and Enhancement via Gaussian Splatting for Novel View Synthesis

Novel view synthesis (NVS) in low-light scenes remains a significant challenge due to degraded inputs characterized by severe noise, low dynamic range (LDR) and unreliable initialization. While recent NeRF-based approaches have shown promising results, most suffer from high computational costs, and some rely on carefully captured or pre-processed data--such as RAW sensor inputs or multi-exposure sequences--which severely limits their practicality. In contrast, 3D Gaussian Splatting (3DGS) enables real-time rendering with competitive visual fidelity; however, existing 3DGS-based methods struggle with low-light sRGB inputs, resulting in unstable Gaussian initialization and ineffective noise suppression. To address these challenges, we propose LL-Gaussian, a novel framework for 3D reconstruction and enhancement from low-light sRGB images, enabling pseudo normal-light novel view synthesis. Our method introduces three key innovations: 1) an end-to-end Low-Light Gaussian Initialization Module (LLGIM) that leverages dense priors from learning-based MVS approach to generate high-quality initial point clouds; 2) a dual-branch Gaussian decomposition model that disentangles intrinsic scene properties (reflectance and illumination) from transient interference, enabling stable and interpretable optimization; 3) an unsupervised optimization strategy guided by both physical constrains and diffusion prior to jointly steer decomposition and enhancement. Additionally, we contribute a challenging dataset collected in extreme low-light environments and demonstrate the effectiveness of LL-Gaussian. Compared to state-of-the-art NeRF-based methods, LL-Gaussian achieves up to 2,000 times faster inference and reduces training time to just 2%, while delivering superior reconstruction and rendering quality.

cs.CV↗

arXiv2025

Small-Cell-Based Fast Active Learning of Machine Learning Interatomic Potentials

Machine learning interatomic potentials (MLIPs) are often trained with on-the-fly active learning, where sampled configurations from atomistic simulations are added to the training set. However, this approach is limited by the high computational cost of ab initio calculations for large systems. Recent works have shown that MLIPs trained on small cells (1-8 atoms) rival the accuracy of large-cell models (100s of atoms) at far lower computational cost. Herein, we refer to these as small-cell and large-cell training, respectively. In this work, we iterate on earlier small-cell training approaches and characterize our resultant small-cell protocol. Potassium and sodium-potassium systems were studied: the former, a simpler system benchmarked in detail; the latter, a more complex binary system for further validation. Our small-cell training approach achieves up to two orders of magnitude of cost savings compared to large-cell (54-atom) training, with some training runs requiring fewer than 120 core-hours. Static and thermodynamic properties predicted using the MLIPs were evaluated, with small-cell training in both systems yielding strong ab initio agreement. Small cells appear to encode the necessary information to model complex large-scale phenomena--solid-liquid interfaces, critical exponents, diverse concentrations--even when the training cells themselves are too small to accommodate these phenomena. Based on these tests, we provide analysis and recommendations.

cond-mat.mtrl-sci↗

arXiv2025

PINP: Physics-Informed Neural Predictor with latent estimation of fluid flows

Accurately predicting fluid dynamics and evolution has been a long-standing challenge in physical sciences. Conventional deep learning methods often rely on the nonlinear modeling capabilities of neural networks to establish mappings between past and future states, overlooking the fluid dynamics, or only modeling the velocity field, neglecting the coupling of multiple physical quantities. In this paper, we propose a new physics-informed learning approach that incorporates coupled physical quantities into the prediction process to assist with forecasting. Central to our method lies in the discretization of physical equations, which are directly integrated into the model architecture and loss function. This integration enables the model to provide robust, long-term future predictions. By incorporating physical equations, our model demonstrates temporal extrapolation and spatial generalization capabilities. Experimental results show that our approach achieves the state-of-the-art performance in spatiotemporal prediction across both numerical simulations and real-world extreme-precipitation nowcasting benchmarks.

cs.LG↗

arXiv2025

Control Map Distribution using Map Query Bank for Online Map Generation

Reliable autonomous driving systems require high-definition (HD) map that contains detailed map information for planning and navigation. However, pre-build HD map requires a large cost. Visual-based Online Map Generation (OMG) has become an alternative low-cost solution to build a local HD map. Query-based BEV Transformer has been a base model for this task. This model learns HD map predictions from an initial map queries distribution which is obtained by offline optimization on training set. Besides the quality of BEV feature, the performance of this model also highly relies on the capacity of initial map query distribution. However, this distribution is limited because the limited query number. To make map predictions optimal on each test sample, it is essential to generate a suitable initial distribution for each specific scenario. This paper proposes to decompose the whole HD map distribution into a set of point representations, namely map query bank (MQBank). To build specific map query initial distributions of different scenarios, low-cost standard definition map (SD map) data is introduced as a kind of prior knowledge. Moreover, each layer of map decoder network learns instance-level map query features, which will lose detailed information of each point. However, BEV feature map is a point-level dense feature. It is important to keep point-level information in map queries when interacting with BEV feature map. This can also be solved with map query bank method. Final experiments show a new insight on SD map prior and a new record on OpenLaneV2 benchmark with 40.5%, 45.7% mAP on vehicle lane and pedestrian area.

cs.CV↗

arXiv2025

MTL-LoRA: Low-Rank Adaptation for Multi-Task Learning

Parameter-efficient fine-tuning (PEFT) has been widely employed for domain adaptation, with LoRA being one of the most prominent methods due to its simplicity and effectiveness. However, in multi-task learning (MTL) scenarios, LoRA tends to obscure the distinction between tasks by projecting sparse high-dimensional features from different tasks into the same dense low-dimensional intrinsic space. This leads to task interference and suboptimal performance for LoRA and its variants. To tackle this challenge, we propose MTL-LoRA, which retains the advantages of low-rank adaptation while significantly enhancing MTL capabilities. MTL-LoRA augments LoRA by incorporating additional task-adaptive parameters that differentiate task-specific information and capture shared knowledge across various tasks within low-dimensional spaces. This approach enables pre-trained models to jointly adapt to different target domains with a limited number of trainable parameters. Comprehensive experimental results, including evaluations on public academic benchmarks for natural language understanding, commonsense reasoning, and image-text understanding, as well as real-world industrial text Ads relevance datasets, demonstrate that MTL-LoRA outperforms LoRA and its various variants with comparable or even fewer learnable parameters in MTL setting.

cs.LG↗

arXiv2025

TITE-STEIN: Time-to-event Simple Toxicity and Efficacy Interval Design to Accelerate Phase I/II Trials

Oncology dose-finding trials are shifting from identifying the maximum tolerated dose (MTD) to determining the optimal biological dose (OBD), driven by the need for efficient methods that consider both toxicity and efficacy. This is particularly important for novel therapies, such as immunotherapies and molecularly targeted therapies, which often exhibit non-monotonic dose-efficacy curves. However, making timely adaptive dosing decisions is challenging due to the rapid patient accrual rate and the late-onset toxicity and/or efficacy outcomes associated with these therapies. The Simple Toxicity and Efficacy Interval (STEIN) design has demonstrated strong performance in accommodating diverse dose-efficacy patterns and incorporating both toxicity and efficacy outcomes to select the OBD. However, the rapid accrual of patients and the often-delayed onset of toxicity and/or efficacy pose challenges to timely adaptive dose decisions. To address these challenges, we propose TITE-STEIN, a model-assisted design that incorporates time-to-event (TITE) outcomes for toxicity and/or efficacy, by extending STEIN. In this article, we demonstrate that TITE-STEIN significantly shortens trial duration compared to STEIN. Furthermore, by integrating an OBD verification procedure during OBD selection, TITE-STEIN effectively mitigates the risk of exposing patients to inadmissible doses when the OBD does not exist. Extensive simulations demonstrate that TITE-STEIN outperforms existing TITE designs, including TITE-BONI12, TITE-BOIN-ET, LO-TC, and Joint TITE-CRM, by selecting the OBD more accurately, allocating more patients to it, and improving overdose control.

stat.ME↗

arXiv2025

Amorphous silicon structures generated using a moment tensor potential and the activation relaxation technique nouveau

Preparing realistic atom-scale models of amorphous silicon (a-Si) is a decades-old condensed matter physics challenge. Herein, we combine the Activation Relaxation Technique nouveau (ARTn) to a Moment Tensor Potential (MTP) to generate seven a-Si models containing between 216 and 4096 atoms. A thorough analysis of their short-range and medium-range structural properties is performed, alongside assessments of excess energy and mechanical properties. The seven ARTn-MTP models are compared with available experimental data and other high quality a-Si models present in the literature. The seven ARTn-MTP a-Si models are in excellent agreement with available experimental data. Notably, several of our models, including the 216-atom, 512-atom, and 1000-atom a-Si models, exhibit low coordination defects without any traces of crystalline grains. Historically overlooked in previous research, our study underlines the need to assess the validity of the continuous random-network hypothesis for the description of perfect amorphous model by characterizing local crystalline environment and to explore the crystallisation process of a-Si through modelling.

cond-mat.dis-nn↗

arXiv2025

FlexWorld: Progressively Expanding 3D Scenes for Flexiable-View Synthesis

Generating flexible-view 3D scenes, including 360° rotation and zooming, from single images is challenging due to a lack of 3D data. To this end, we introduce FlexWorld, a novel framework consisting of two key components: (1) a strong video-to-video (V2V) diffusion model to generate high-quality novel view images from incomplete input rendered from a coarse scene, and (2) a progressive expansion process to construct a complete 3D scene. In particular, leveraging an advanced pre-trained video model and accurate depth-estimated training pairs, our V2V model can generate novel views under large camera pose variations. Building upon it, FlexWorld progressively generates new 3D content and integrates it into the global scene through geometry-aware scene fusion. Extensive experiments demonstrate the effectiveness of FlexWorld in generating high-quality novel view videos and flexible-view 3D scenes from single images, achieving superior visual quality under multiple popular metrics and datasets compared to existing state-of-the-art methods. Qualitatively, we highlight that FlexWorld can generate high-fidelity scenes with flexible views like 360° rotations and zooming. Project page: https://ml-gsai.github.io/FlexWorld.

cs.CV↗

arXiv2025

A Token-level Text Image Foundation Model for Document Understanding

In recent years, general visual foundation models (VFMs) have witnessed increasing adoption, particularly as image encoders for popular multi-modal large language models (MLLMs). However, without semantically fine-grained supervision, these models still encounter fundamental prediction errors in the context of downstream text-image-related tasks, i.e., perception, understanding and reasoning with images containing small and dense texts. To bridge this gap, we develop TokenOCR, the first token-level visual foundation model specifically tailored for text-image-related tasks, designed to support a variety of traditional downstream applications. To facilitate the pretraining of TokenOCR, we also devise a high-quality data production pipeline that constructs the first token-level image text dataset, TokenIT, comprising 20 million images and 1.8 billion token-mask pairs. Furthermore, leveraging this foundation with exceptional image-as-text capability, we seamlessly replace previous VFMs with TokenOCR to construct a document-level MLLM, TokenVL, for VQA-based document understanding tasks. Finally, extensive experiments demonstrate the effectiveness of TokenOCR and TokenVL. Code, datasets, and weights will be available at https://github.com/Token-family/TokenFD.

cs.CV↗

arXiv2025

OmniEraser: Remove Objects and Their Effects in Images with Paired Video-Frame Data

Inpainting algorithms have achieved remarkable progress in removing objects from images, yet still face two challenges: 1) struggle to handle the object's visual effects such as shadow and reflection; 2) easily generate shape-like artifacts and unintended content. In this paper, we propose Video4Removal, a large-scale dataset comprising over 100,000 high-quality samples with realistic object shadows and reflections. By constructing object-background pairs from video frames with off-the-shelf vision models, the labor costs of data acquisition can be significantly reduced. To avoid generating shape-like artifacts and unintended content, we propose Object-Background Guidance, an elaborated paradigm that takes both the foreground object and background images. It can guide the diffusion process to harness richer contextual information. Based on the above two designs, we present OmniEraser, a novel method that seamlessly removes objects and their visual effects using only object masks as input. Extensive experiments show that OmniEraser significantly outperforms previous methods, particularly in complex in-the-wild scenes. And it also exhibits a strong generalization ability in anime-style images. Datasets, models, and codes will be published.

cs.CV↗