arXiv subjects

Ze Li

Publications and source records attributed to Ze Li.

At least 19 recordsLinked to original sources

arXiv2026

Inverse Heat Source Problems from Boundary Flux and Interior Observations on Sets of Low Hausdorff Dimension

This paper investigates conditional stability for inverse source problems for the heat equation with a known temporal factor and an unknown spatial component in a bounded $C^{1,1}$ domain. We focus on observations supported on sets of low Hausdorff dimension and establish conditional stability in this setting. For boundary observations on compact sets of positive $q$-dimensional Hausdorff content, we establish logarithmic stability from full-time boundary flux observations and double-logarithmic stability from delayed-time boundary flux observations. The admissible dimensional ranges are $q>d-2$ when the observation set is contained in a flat boundary patch and $q>d-1-c_{d+1}$ on a general $C^{1,1}$ boundary, where $c_{d+1}>0$ depends only on the dimension. A key ingredient in deriving these results is a new boundary spectral inequality for the Dirichlet Laplacian, which controls a finite Dirichlet spectral sum through observations of the normal derivative of its elliptic extension on such a boundary set. Our results also cover inverse heat source problems with interior observations on sets of positive $q$-dimensional Hausdorff content for some $q>d-1$, yielding logarithmic stability from full-time observations for general sources in $H_0^1(\Omega)$ and H\"older stability from terminal-time observations for sources in a suitable spectral Gevrey class.

math.SP↗

arXiv2026

Indirect evidence of the $2175\,\mathring{\mathrm{A}}$ extinction bump within the dusty torus of SDSS J141945.50+524648.0

We present a multi-wavelength study of the quasar SDSS J141945.50$+$524648.0 ($z=1.1599$), a member of the newly identified population of quasar-associated $2175\,\mathring{\mathrm{A}}$ dust absorbers. Utilizing JWST/NIRCam observations from the SAPPHIRES survey and archival data, we analyze the prominent $2175\,\mathring{\mathrm{A}}$ extinction bump detected in this object. The bump parameters are highly consistent with the Milky Way extinction curve, implying similar dust properties. Spectral analysis reveals a $\mathrm{Mg\,II}$ absorption doublet near the systemic velocity. Joint fitting of the $\mathrm{Mg\,II}$ and $\mathrm{Fe\,II}$ absorption lines favors a partial-covering model, yielding $C_{f,\rm BEL}=0.11^{+0.12}_{-0.07}$. Together with the small velocity offset ($\Delta v\approx96~{\rm km\,s^{-1}}$), these results favor an intrinsic absorber rather than an intervening system. Infrared SED modeling and double-peaked $\mathrm{Pa\,\alpha}$ emission indicate consistent orientations of the dusty torus ($\theta\approx60^{+7}_{-8}$ deg) and accretion disk ($i=63.7^{+10.9}_{-7.8}$ deg), suggesting a rim-penetrating line of sight through the torus. This configuration naturally explains the infrared emission, continuum extinction, and associated $2175\,\mathring{\mathrm{A}}$ bump, although a contribution from the host-galaxy ISM cannot be excluded. If associated with the torus, the carbonaceous carriers of the $2175\,\mathring{\mathrm{A}}$ feature may survive the intense AGN radiation field through localized shielding in optically thick dusty clumps. These results highlight the role of viewing geometry and dust distribution in regulating dust survival in quasar environments.

astro-ph.GA↗

arXiv2026

Hausdorff type Time-Trace Observability for Airy Equations on the Line and Point Observability on the Torus

The main results of this paper are threefold. First, we prove an observability inequality for the Airy equation on the real line from Hausdorff-thick sets in a time-trace sense for every observation time $T>0$. The observation functional is a block supremum of $L^2(0,T)$ time traces over the Hausdorff-thick set. Second, we prove observability inequalities for the Airy equation on the real line with observations on some periodic sets, which in particular yields observability on a class of spatial point sequences. Third, we give a necessary and sufficient condition for finite point observability for the Airy equation on the torus with a bounded real-valued potential. Indeed, for a finite observation set $F$, a Kalman rank condition on a finite-dimensional invariant subspace is found. As a corollary, we obtain sharp point observability results for the Airy and linear KdV equations on the torus. Moreover, for potentials with finite order regularity assumptions we prove that every observation set with an accumulation point, in particular any set of positive Hausdorff dimension, gives an observability inequality for each observation time $T>0$. Since the Airy equation has neither high frequency exponential decay nor pointwise smoothing effects, which are essential in recent works on Hausdorff type observation results on heat equations, we introduce several new ideas adapted to the Airy case.

math.AP↗

arXiv2026

ActionNex: A Virtual Outage Manager for Cloud Computing

Outage management in large-scale cloud operations remains heavily manual, requiring rapid triage, cross-team coordination, and experience-driven decisions under partial observability. We present \textbf{ActionNex}, a production-grade agentic system that supports end-to-end outage assistance, including real-time updates, knowledge distillation, and role- and stage-conditioned next-best action recommendations. ActionNex ingests multimodal operational signals (e.g., outage content, telemetry, and human communications) and compresses them into critical events that represent meaningful state transitions. It couples this perception layer with a hierarchical memory subsystem: long-term Key-Condition-Action (KCA) knowledge distilled from playbooks and historical executions, episodic memory of prior outages, and working memory of the live context. A reasoning agent aligns current critical events to preconditions, retrieves relevant memories, and generates actionable recommendations; executed human actions serve as an implicit feedback signal to enable continual self-evolution in a human-agent hybrid system. We evaluate ActionNex on eight real Azure outages (8M tokens, 4,000 critical events) using two complementary ground-truth action sets, achieving 71.4\% precision and 52.8-54.8\% recall. The system has been piloted in production and has received positive early feedback.

cs.AI↗

arXiv2026

Language-Invariant Multilingual Speaker Verification for the TidyVoice 2026 Challenge

Multilingual speaker verification (SV) remains challenging due to limited cross-lingual data and language-dependent information in speaker embeddings. This paper presents a language-invariant multilingual SV system for the TidyVoice 2026 Challenge. We adopt the multilingual self-supervised w2v-BERT 2.0 model as the backbone, enhanced with Layer Adapters and Multi-scale Feature Aggregation to better exploit multi-layer representations. A language-adversarial training strategy with a Gradient Reversal Layer is applied to promote language-invariant speaker embeddings. Moreover, a multilingual zero-shot text-to-speech system is used to synthesize speech in multiple languages, improving language diversity. Experimental results demonstrate that fine-tuning the large-scale pretrained model yields competitive performance, while language-adversarial training further enhances robustness. In addition, synthetic speech augmentation provides additional gains under limited training data conditions. Source code is available at https://github.com/ZXHY-82/LI-MSV-TidyVoice2026.

eess.AS↗

arXiv2026

Fast Low-light Enhancement and Deblurring for 3D Dark Scenes

Novel view synthesis from low-light, noisy, and motion-blurred imagery remains a valuable and challenging task. Current volumetric rendering methods struggle with compound degradation, and sequential 2D preprocessing introduces artifacts due to interdependencies. In this work, we introduce FLED-GS, a fast low-light enhancement and deblurring framework that reformulates 3D scene restoration as an alternating cycle of enhancement and reconstruction. Specifically, FLED-GS inserts several intermediate brightness anchors to enable progressive recovery, preventing noise blow-up from harming deblurring or geometry. Each iteration sharpens inputs with an off-the-shelf 2D deblurrer and then performs noise-aware 3DGS reconstruction that estimates and suppresses noise while producing clean priors for the next level. Experiments show FLED-GS outperforms state-of-the-art LuSh-NeRF, achieving 21$\times$ faster training and 11$\times$ faster rendering.

cs.CV↗

arXiv2026

Quantization-scheme-Independent Energy and Its Implications for Holographic Bounds

In holographic duality, the total energy of the dual field theory is obtained from the holographic renormalization, which depends not only on the bulk geometry but also on the choice of quantization schemes. We point out that the validity of several widely studied holographic inequalities -- including the AdS Penrose inequality, the late-time bound on entanglement entropy growth, and the growth-rate limits of CV and CA complexities -- depends on the choice of quantization schemes. Motivated by this issue, we introduce a modified total energy, which is still computed via holographic renormalization but the final value is independent of the choice of quantization schemes. We verify that this modified energy removes the apparent violations of these bounds that arise from quantization-scheme dependence in the model of massive scalar field. Our results suggest that our modified total energy provides a more robust notion of energy when we talk about above inequalities in holographic settings.

hep-th↗

arXiv2025

Enhancing Speaker Verification with w2v-BERT 2.0 and Knowledge Distillation guided Structured Pruning

Large-scale self-supervised Pre-Trained Models (PTMs) have shown significant improvements in the speaker verification (SV) task by providing rich feature representations. In this paper, we utilize w2v-BERT 2.0, a model with approximately 600 million parameters trained on 4.5 million hours of unlabeled data across 143 languages, for the SV task. The MFA structure with Layer Adapter is employed to process the multi-layer feature outputs from the PTM and extract speaker embeddings. Additionally, we incorporate LoRA for efficient fine-tuning. Our model achieves state-of-the-art results with 0.12% and 0.55% EER on the Vox1-O and Vox1-H test sets, respectively. Furthermore, we apply knowledge distillation guided structured pruning, reducing the model size by 80% while achieving only a 0.04% EER degradation. Source code and models are released at https://github.com/ZXHY-82/w2v-BERT-2.0_SV.

eess.AS↗

arXiv2025

Desclarizing the Wormhole to Black Hole with Negative Mass

We construct new black holes using the curvature-induced scalarization mechanism in Einstein gravity coupled to a massless phantom scalar. We find that the general wormhole solutions with independent scalar charge become descalarized so that the resulting black hole scalar hair becomes secondary and a function of the mass only. The long-range force between the two identical black holes can be attractive, zero or repulsive, depending on the mass/scalar charge relation. Furthermore, the black hole mass can be negative. Our finding suggests that exotic matter responsible for wormholes can also lead to exotic black holes.

gr-qc↗

arXiv2025

Charged Taub-NUT type Black Holes in Einstein-Weyl-Maxwell Theory

We numerically construct new charged Taub-NUT-type black hole solutions in four-dimensional Einstein-Weyl-Maxwell theory. We explore the effects of the NUT parameter and the electric charge parameter on the black hole solutions in great detail. Compared with static spherical black holes in the same theory, there is one major difference: there are three branches of NUT solutions, whereas there are only two branches of static spherical black hole solutions. Compared with neutral NUT solutions in Weyl gravity, where the solutions intersect with each other, the charged NUT solutions we find are disconnected.

gr-qc↗

arXiv2025

The DKU System for Multi-Speaker Automatic Speech Recognition in MLC-SLM Challenge

We present the DKU system for Task 2 of the MLC-SLM Challenge, which aims to perform multi-speaker automatic speech recognition directly from raw audio without Oracle speaker labels or time boundaries. Our approach builds upon a diarization-aware framework integrating speaker embeddings and temporal utterance boundaries into a Qwen2.5-based large language model (LLM). Then, we enhance the system's multilingual performance by fine-tuning language-specific adapters and LoRA modules within the LLM decoder. Finally, our system achieves the tcpWER of 23.56\% and 18.08\% on the development and test sets of the MLC-SLM dataset, substantially outperforming the official baseline.

eess.AS↗

arXiv2025

Robust Low-light Scene Restoration via Illumination Transition

Synthesizing normal-light novel views from low-light multiview images is an important yet challenging task, given the low visibility and high ISO noise present in the input images. Existing low-light enhancement methods often struggle to effectively preprocess such low-light inputs, as they fail to consider correlations among multiple views. Although other state-of-the-art methods have introduced illumination-related components offering alternative solutions to the problem, they often result in drawbacks such as color distortions and artifacts, and they provide limited denoising effectiveness. In this paper, we propose a novel Robust Low-light Scene Restoration framework (RoSe), which enables effective synthesis of novel views in normal lighting conditions from low-light multiview image inputs, by formulating the task as an illuminance transition estimation problem in 3D space, conceptualizing it as a specialized rendering task. This multiview-consistent illuminance transition field establishes a robust connection between low-light and normal-light conditions. By further exploiting the inherent low-rank property of illumination to constrain the transition representation, we achieve more effective denoising without complex 2D techniques or explicit noise modeling. To implement RoSe, we design a concise dual-branch architecture and introduce a low-rank denoising module. Experiments demonstrate that RoSe significantly outperforms state-of-the-art models in both rendering quality and multiview consistency on standard benchmarks. The codes and data are available at https://pegasus2004.github.io/RoSe.

cs.CV↗

arXiv2025

EvolvTrip: Enhancing Literary Character Understanding with Temporal Theory-of-Mind Graphs

A compelling portrayal of characters is essential to the success of narrative writing. For readers, appreciating a character's traits requires the ability to infer their evolving beliefs, desires, and intentions over the course of a complex storyline, a cognitive skill known as Theory-of-Mind (ToM). Performing ToM reasoning in prolonged narratives requires readers to integrate historical context with current narrative information, a task at which humans excel but Large Language Models (LLMs) often struggle. To systematically evaluate LLMs' ToM reasoning capability in long narratives, we construct LitCharToM, a benchmark of character-centric questions across four ToM dimensions from classic literature. Further, we introduce EvolvTrip, a perspective-aware temporal knowledge graph that tracks psychological development throughout narratives. Our experiments demonstrate that EvolvTrip consistently enhances performance of LLMs across varying scales, even in challenging extended-context scenarios. EvolvTrip proves to be particularly valuable for smaller models, partially bridging the performance gap with larger LLMs and showing great compatibility with lengthy narratives. Our findings highlight the importance of explicit representation of temporal character mental states in narrative comprehension and offer a foundation for more sophisticated character understanding. Our data and code are publicly available at https://github.com/Bernard-Yang/EvolvTrip.

cs.CL↗

arXiv2025

Diarization-Aware Multi-Speaker Automatic Speech Recognition via Large Language Models

Multi-speaker automatic speech recognition (MS-ASR) faces significant challenges in transcribing overlapped speech, a task critical for applications like meeting transcription and conversational analysis. While serialized output training (SOT)-style methods serve as common solutions, they often discard absolute timing information, limiting their utility in time-sensitive scenarios. Leveraging recent advances in large language models (LLMs) for conversational audio processing, we propose a novel diarization-aware multi-speaker ASR system that integrates speaker diarization with LLM-based transcription. Our framework processes structured diarization inputs alongside frame-level speaker and semantic embeddings, enabling the LLM to generate segment-level transcriptions. Experiments demonstrate that the system achieves robust performance in multilingual dyadic conversations and excels in complex, high-overlap multi-speaker meeting scenarios. This work highlights the potential of LLMs as unified back-ends for joint speaker-aware segmentation and transcription.

eess.AS↗

arXiv2025

On Charged Black Holes in Einstein-Weyl-Maxwell Theory

There exist two branches of static and spherically symmetric black hole solutions in Einstein-Weyl theory: one is the Schwarzschild black hole, and the other is a numerically constructed black hole that bifurcates from the Schwarzschild solution. Similarly, there are two branches of charged black holes in Einstein-Weyl-Maxwell theory. We have uncovered the relationships between the charged black holes and the neutral ones. The two charged black holes branch out from the bifurcation point of the neutral ones once charge is added. We found that one of the charged black holes is entirely different from the Reissner-Nordstr\"om (RN) black hole, while the other is similar to the RN black hole. In particular, the RN-like black hole approaches the RN black hole as the charge increases. We calculated the charge-to-mass ratio of the RN-like charged black hole in the near-extremal limit, and the value is less than that of the extremal RN black hole. Since we consider the higher-derivative Weyl square term as part of the classical gravity theory, rather than as a quantum effect, our result sets a lower bound on the charge-to-mass ratio in the context of the Weak Gravity Conjecture.

gr-qc↗

arXiv2025

An abstract uncertainty principle with applications

Under Wigdersons' framework and by sorting out the technical points in the recent works of Tang (J. Fourier Anal. Appl. 31 (2025)) and Dias-Luef-Prata (J. Math. Pures Appl. (9) 198 (2025)), we prove an abstract uncertainty principle for functions in the $L^p$ setting. An immediate consequence is a new uncertainty principle for the Fourier transform, unifying and extending many existing results. More applications are shown for PDEs, including the moment growth estimates for some linear and nonlinear dispersive equations, and a type of weighted lower bound estimate for the spacetime moment of the Schr\"{o}dinger equation and heat equation inspired by the control theory.

math.AP↗

arXiv2025

Robust Low-Light Human Pose Estimation through Illumination-Texture Modulation

As critical visual details become obscured, the low visibility and high ISO noise in extremely low-light images pose a significant challenge to human pose estimation. Current methods fail to provide high-quality representations due to reliance on pixel-level enhancements that compromise semantics and the inability to effectively handle extreme low-light conditions for robust feature learning. In this work, we propose a frequency-based framework for low-light human pose estimation, rooted in the "divide-and-conquer" principle. Instead of uniformly enhancing the entire image, our method focuses on task-relevant information. By applying dynamic illumination correction to the low-frequency components and low-rank denoising to the high-frequency components, we effectively enhance both the semantic and texture information essential for accurate pose estimation. As a result, this targeted enhancement method results in robust, high-quality representations, significantly improving pose estimation performance. Extensive experiments demonstrating its superiority over state-of-the-art methods in various challenging low-light scenarios.

cs.CV↗

arXiv2024

Revisiting holographic dark energy from the perspective of multi-messenger gravitational wave astronomy: future joint observations with short gamma-ray bursts

The advent of third-generation (3G) gravitational-wave (GW) detectors opens new opportunities for multi-messenger observations of binary neutron star merger events, holding significant potential for probing the history of cosmic expansion. In this paper, we investigate the holographic dark energy (HDE) model by using the future GW standard siren data observed from the 3G GW detectors and the short $\gamma$-ray burst THESEUS-like detector joint observations. We find that GW data alone can achieve a relatively precise estimation of the Hubble constant, with precision of $0.2\%$-$0.6\%$, but its ability to constrain other cosmological parameters remains limited. Nonetheless, since the GW data can break parameter degeneracies generated by the mainstream EM observations, CMB + BAO + SN (CBS), GW standard sirens play a crucial role in enhancing the accuracy of parameter estimation. With the addition of GW data to CBS, the constraints on cosmological parameters $H_0$, $c$ and $\Omega_{\rm{m}}$ can be improved by $63\%$-$88\%$, $27\%$-$44\%$ and $55\%$-$70\%$. In summary, observations of GW standard sirens from 3G GW detectors could be pivotal in helping solve the Hubble tension and probe the fundamental nature of dark energy.

astro-ph.CO↗