arXiv subjects

Fudong Wang

Publications and source records attributed to Fudong Wang.

At least 19 recordsLinked to original sources

arXiv2026

Slow-light-enhanced Atomic Frequency Comb Quantum Memory in Stoichiometric EuCl$_3 \cdot$ 6D$_2$O

Rare-earth-doped crystals are promising candidates for quantum storage, yet their performance in free-space configurations is fundamentally restricted by low optical depth. Here, we demonstrate high-efficiency quantum storage in a stoichiometric EuCl$_3 \cdot$ 6D$_2$O crystal, which intrinsically provides high optical density without the complexity of cavity implementation. We show that in this high-density regime, the system exhibits significant slow-light-like effects, including dispersion-induced echo delays and finesse-dependent echo intensity modulation. We develop a unified theoretical framework showing how absorption and dispersion work in concert to mediate echo generation. We achieve storage efficiencies of 42.9% for classical light and 34.4% for weak coherent pulses, alongside 90% efficiency for slow-light storage. These findings validate EuCl$_3 \cdot$ 6D$_2$O as a robust platform, establishing a viable pathway for scalable solid-state quantum memory.

quant-ph↗

arXiv2026

Realization of Backward Retrieval in a Stark-modulated Spin-wave Quantum Memory

We report the first experimental realization of backward retrieval in a spin-wave quantum memory based on a Stark-echo-modulated protocol in Eu3+:Y2SiO5. By using Stark control, we preserve the full optical depth of the ensemble while suppressing coherent noise, enabling conditional storage fidelities above 97%. Our analysis shows that the present backward-retrieval efficiency is mainly limited by technical imperfections rather than by fundamental constraints. With realistic engineering improvements, backward retrieval in this protocol could move beyond the reabsorption-limited forward-emission regime. The protocol is also compatible with cavity-enhanced operation, offering an additional route toward higher efficiencies. These findings establish Stark-echo modulation as a practical and scalable route to high-efficiency, long-lived solid-state quantum memories.

quant-ph↗

arXiv2025

Ming-Flash-Omni: A Sparse, Unified Architecture for Multimodal Perception and Generation

We propose Ming-Flash-Omni, an upgraded version of Ming-Omni, built upon a sparser Mixture-of-Experts (MoE) variant of Ling-Flash-2.0 with 100 billion total parameters, of which only 6.1 billion are active per token. This architecture enables highly efficient scaling (dramatically improving computational efficiency while significantly expanding model capacity) and empowers stronger unified multimodal intelligence across vision, speech, and language, representing a key step toward Artificial General Intelligence (AGI). Compared to its predecessor, the upgraded version exhibits substantial improvements across multimodal understanding and generation. Notably, it achieves strong performance on vision-language understanding benchmarks, with overall scores on par with Gemini 2.5 Pro, and enables seamless switching among multimodal tasks in multi-turn interactions. In speech, it achieves strong performance in contextual and dialect-aware ASR while enabling joint, continuous-generation of speech, sound, and music. In vision, it introduces generative semantic segmentation that achieves competitive standalone performance and enhances spatial control and editing consistency, alongside marked improvements in identity preservation, and high-fidelity in-image text rendering. Together, these capabilities demonstrate that a single unified model can serve as a practical foundation for general-purpose multimodal intelligence.

cs.CV↗

arXiv2025

M2-Reasoning: Empowering MLLMs with Unified General and Spatial Reasoning

Recent advancements in Multimodal Large Language Models (MLLMs), particularly through Reinforcement Learning with Verifiable Rewards (RLVR), have significantly enhanced their reasoning abilities. However, a critical gap persists: these models struggle with dynamic spatial interactions, a capability essential for real-world applications. To bridge this gap, we introduce M2-Reasoning-7B, a model designed to excel in both general and spatial reasoning. Our approach integrates two key innovations: (1) a novel data pipeline that generates 294.2K high-quality data samples (168K for cold-start fine-tuning and 126.2K for RLVR), which feature logically coherent reasoning trajectories and have undergone comprehensive assessment; and (2) a dynamic multi-task training strategy with step-wise optimization to mitigate conflicts between data, and task-specific rewards for delivering tailored incentive signals. This combination of curated data and advanced training allows M2-Reasoning-7B to set a new state-of-the-art (SOTA) across 8 benchmarks, showcasing superior performance in both general and spatial reasoning domains.

cs.AI↗

arXiv2025

LangScene-X: Reconstruct Generalizable 3D Language-Embedded Scenes with TriMap Video Diffusion

Recovering 3D structures with open-vocabulary scene understanding from 2D images is a fundamental but daunting task. Recent developments have achieved this by performing per-scene optimization with embedded language information. However, they heavily rely on the calibrated dense-view reconstruction paradigm, thereby suffering from severe rendering artifacts and implausible semantic synthesis when limited views are available. In this paper, we introduce a novel generative framework, coined LangScene-X, to unify and generate 3D consistent multi-modality information for reconstruction and understanding. Powered by the generative capability of creating more consistent novel observations, we can build generalizable 3D language-embedded scenes from only sparse views. Specifically, we first train a TriMap video diffusion model that can generate appearance (RGBs), geometry (normals), and semantics (segmentation maps) from sparse inputs through progressive knowledge integration. Furthermore, we propose a Language Quantized Compressor (LQC), trained on large-scale image datasets, to efficiently encode language embeddings, enabling cross-scene generalization without per-scene retraining. Finally, we reconstruct the language surface fields by aligning language information onto the surface of 3D scenes, enabling open-ended language queries. Extensive experiments on real-world data demonstrate the superiority of our LangScene-X over state-of-the-art methods in terms of quality and generalizability. Project Page: https://liuff19.github.io/LangScene-X.

cs.CV↗

arXiv2025

Ming-Omni: A Unified Multimodal Model for Perception and Generation

We propose Ming-Omni, a unified multimodal model capable of processing images, text, audio, and video, while demonstrating strong proficiency in both speech and image generation. Ming-Omni employs dedicated encoders to extract tokens from different modalities, which are then processed by Ling, an MoE architecture equipped with newly proposed modality-specific routers. This design enables a single model to efficiently process and fuse multimodal inputs within a unified framework, thereby facilitating diverse tasks without requiring separate models, task-specific fine-tuning, or structural redesign. Importantly, Ming-Omni extends beyond conventional multimodal models by supporting audio and image generation. This is achieved through the integration of an advanced audio decoder for natural-sounding speech and Ming-Lite-Uni for high-quality image generation, which also allow the model to engage in context-aware chatting, perform text-to-speech conversion, and conduct versatile image editing. Our experimental results showcase Ming-Omni offers a powerful solution for unified perception and generation across all modalities. Notably, our proposed Ming-Omni is the first open-source model we are aware of to match GPT-4o in modality support, and we release all code and model weights to encourage further research and development in the community.

cs.AI↗

arXiv2025

Spectral theory of soliton gases for the defocusing NLS equation

In this paper we derive Nonlinear Dispersion Relations (NDR) for the defocusing NLS (dark) soliton gas using the idea of thermodynamic limit of quasimomentum and quasienergy differentials on the underlying family of Riemann surfaces. It turns out that the obtained NDR are closely connected with the recently studied NDR for circular soliton gas for the focusing NLS. We find solutions for the kinetic equation for the defocusing NLS soliton condensate, which is defined by the endpoints of the spectral support $\Gamma$ for the NDR. It turns out that, similarly to KdV soliton condensates (\cite{CERT}), the evolution of these endpoints is governed by the defocusing NLS-Whitham equations (\cite{Kodama}). We also study the Riemann problem for step initial data and the kurtosis of genus zero and one defNLS condensates, where we proved that the kurtosis of genus one condensate can not exceed 3/2, whereas for genus zero condensate the kurtosis is always 1.

nlin.PS↗

arXiv2025

Ultralong-lived Coherent States in Eu$^{3+}$:Y$_2$O$_3$ Optical Ceramics for Quantum Memories

Rare earth ions (REI) in solid materials are among the leading systems for quantum technology applications. However, developing practical REI quantum devices with long-lived coherent states remains challenging due to great growth difficulties of high-quality REI materials and a lack of comprehensive understanding of REI's decoherence mechanisms. Here we realize a record optical coherence time of 421.5 $\pm$ 10.5 $\mu$s for the $^7$F$_0\rightarrow^5$D$_0$ transition and more than 30 hours lifetime for the $^7$F$_0$ hyperfine spin states in Eu$^{3+}$:Y$_2$O$_3$ optical ceramics. We report the elimination of two-level-system induced optical decoherence in short-range ordered crystals. Meanwhile, a new decoherence mechanism caused by new kinds of perturbing magnetic centers is identified below 1.5 K. We further demonstrate the coherent light storage over 5 $\mu$s by using the atomic frequency comb protocol. These results open up prospects for the realization of practical quantum memories and large scale quantum communications with REI optical ceramics.

quant-ph↗

arXiv2025

Optical coherence and hyperfine structure of the 7F0-5D0 transition in EuCaWO4

Rare-earth ions doped in crystals with low nuclear-spin densities are highly promising candidates for quantum technology applications. In this study, we investigated the spectroscopic properties of the 7F0 - 5 D0 optical and the hyperfine transitions of Eu3+ ions in a CaWO4 crystal, where the nuclear spin arises solely from the 183W isotope, with a natural abundance of 14%. At a temperature of 3 K, we experimentally identified four distinct crystal field environments for Eu3+ ions in a 0.1 at.% Eu3+ doped CaWO4 crystal. The optical coherence properties of Eu3+ ions in these environments were characterized. Additionally, we resolved the hyperfine structures in the 7F0 ground state and 5D0 excited state, and determined the 7F0 ground state lifetimes using spectral hole-burning techniques. These findings highlight the significant potential of Eu3+:CaWO4 for optical quantum memory applications.

cond-mat.mtrl-sci↗

arXiv2024

GraphicsDreamer: Image to 3D Generation with Physical Consistency

Recently, the surge of efficient and automated 3D AI-generated content (AIGC) methods has increasingly illuminated the path of transforming human imagination into complex 3D structures. However, the automated generation of 3D content is still significantly lags in industrial application. This gap exists because 3D modeling demands high-quality assets with sharp geometry, exquisite topology, and physically based rendering (PBR), among other criteria. To narrow the disparity between generated results and artists' expectations, we introduce GraphicsDreamer, a method for creating highly usable 3D meshes from single images. To better capture the geometry and material details, we integrate the PBR lighting equation into our cross-domain diffusion model, concurrently predicting multi-view color, normal, depth images, and PBR materials. In the geometry fusion stage, we continue to enforce the PBR constraints, ensuring that the generated 3D objects possess reliable texture details, supporting realistic relighting. Furthermore, our method incorporates topology optimization and fast UV unwrapping capabilities, allowing the 3D products to be seamlessly imported into graphics engines. Extensive experiments demonstrate that our model can produce high quality 3D assets in a reasonable time cost compared to previous methods.

cs.GR↗

arXiv2024

Soliton Condensates for the Focusing Nonlinear Schrödinger Equation: a Non-Bound State Case

In this paper, we study the spectral theory of soliton condensates - a special limit of soliton gases - for the focusing NLS (fNLS). In particular, we analyze the kinetic equation for the fNLS circular condensate, which represents the first example of an explicitly solvable fNLS condensate with nontrivial large scale space-time dynamics. Solution of the kinetic equation was obtained by reducing it to Whitham type equations for the endpoints of spectral arcs. We also study the rarefaction and dispersive shock waves for circular condensates, as well as calculate the corresponding average conserved quantities and the kurtosis. We want to note that one of the main objects of the spectral theory - the nonlinear dispersion relations - is introduced in the paper as some special large genus (thermodynamic) limit the Riemann bilinear identities that involve the quasimomentum and the quasienergy meromorphic differentials.

nlin.PS↗

arXiv2022

Recent developments in spectral theory of the focusing NLS soliton and breather gases: the thermodynamic limit of average densities, fluxes and certain meromorphic differentials; periodic gases

In this paper we consider soliton and breather gases for one dimensional integrable focusing Nonlinear Schrödinger Equation (fNLS). We derive average densities and fluxes for such gases by studying the thermodynamic limit of the fNLS finite gap solutions. Thermodynamic limits of quasimomentum, quasienergy and their connections with the corresponding $g$-functions were also established. We then introduce the notion of periodic fNLS gases and calculate for them the average densities, fluxes and thermodynamic limits of meromorphic differentials. Certain accuracy estimates of the obtained results are also included. Our results constitute another step towards the mathematical foundation for the spectral theory of fNLS soliton and breather gases that appeared in work of G. El and A. Tovbis, Phys. Rev. E, 2020.

nlin.PS↗

arXiv2021

A Note on Electrified Droplets

We give an in-depth analysis of a 1-parameter family of electrified droplets first described in D. Khavinson et. al. (2005). We also investigate a technique for searching for new solutions to the droplet equation, and rederive via this technique a 1-parameter family of physical droplets, which were first discovered by D. Crowdy (1999). We speculate on extensions of these solutions, in particular to the case of a droplet with multiple connected components.

math.CV↗

arXiv2021

Learning Canonical View Representation for 3D Shape Recognition with Arbitrary Views

In this paper, we focus on recognizing 3D shapes from arbitrary views, i.e., arbitrary numbers and positions of viewpoints. It is a challenging and realistic setting for view-based 3D shape recognition. We propose a canonical view representation to tackle this challenge. We first transform the original features of arbitrary views to a fixed number of view features, dubbed canonical view representation, by aligning the arbitrary view features to a set of learnable reference view features using optimal transport. In this way, each 3D shape with arbitrary views is represented by a fixed number of canonical view features, which are further aggregated to generate a rich and robust 3D shape representation for shape recognition. We also propose a canonical view feature separation constraint to enforce that the view features in canonical view representation can be embedded into scattered points in a Euclidean space. Experiments on the ModelNet40, ScanObjectNN, and RGBD datasets show that our method achieves competitive results under the fixed viewpoint settings, and significantly outperforms the applicable methods under the arbitrary view setting.

cs.CV↗

arXiv2021

A $\dbar$-steepest descent method for oscillatory Riemann-Hilbert problems

We study the asymptotic behavior of Riemann-Hilbert problems (RHP) arising in the AKNS hierarchy of integrable equations. Our analysis is based on the $\dbar$-steepest descent method. We consider RHPs arising from the inverse scattering transform of the AKNS hierarchy with $H^{1,1}(\R)$ initial data. The analysis will be divided into three regions: fast decay region, oscillating region and self-similarity region (the Painlevé region). The resulting formulas can be directly applied to study the long-time asymptotic of the solutions of integrable equations such as NLS, mKdV and their higher-order generalizations.

nlin.SI↗

arXiv2021

Deep Graph Matching under Quadratic Constraint

Recently, deep learning based methods have demonstrated promising results on the graph matching problem, by relying on the descriptive capability of deep features extracted on graph nodes. However, one main limitation with existing deep graph matching (DGM) methods lies in their ignorance of explicit constraint of graph structures, which may lead the model to be trapped into local minimum in training. In this paper, we propose to explicitly formulate pairwise graph structures as a \textbf{quadratic constraint} incorporated into the DGM framework. The quadratic constraint minimizes the pairwise structural discrepancy between graphs, which can reduce the ambiguities brought by only using the extracted CNN features. Moreover, we present a differentiable implementation to the quadratic constrained-optimization such that it is compatible with the unconstrained deep learning optimizer. To give more precise and proper supervision, a well-designed false matching loss against class imbalance is proposed, which can better penalize the false negatives and false positives with less overfitting. Exhaustive experiments demonstrate that our method competitive performance on real-world datasets.

cs.CV↗

arXiv2020

Zero-Assignment Constraint for Graph Matching with Outliers

Graph matching (GM), as a longstanding problem in computer vision and pattern recognition, still suffers from numerous cluttered outliers in practical applications. To address this issue, we present the zero-assignment constraint (ZAC) for approaching the graph matching problem in the presence of outliers. The underlying idea is to suppress the matchings of outliers by assigning zero-valued vectors to the potential outliers in the obtained optimal correspondence matrix. We provide elaborate theoretical analysis to the problem, i.e., GM with ZAC, and figure out that the GM problem with and without outliers are intrinsically different, which enables us to put forward a sufficient condition to construct valid and reasonable objective function. Consequently, we design an efficient outlier-robust algorithm to significantly reduce the incorrect or redundant matchings caused by numerous outliers. Extensive experiments demonstrate that our method can achieve the state-of-the-art performance in terms of accuracy and efficiency, especially in the presence of numerous outliers.

cs.CV↗

arXiv2019

Landau-Zener-Stückelberg Interferometry in $\mathcal{PT}$-symmetric Non-Hermitian models

We systematically investigate the non-Hermitian generalisations of the Landau-Zener (LZ) transition and the Landau-Zener-Stückelberg (LZS) interferometry. The LZ transition probabilities, or band populations, are calculated for a generic non-Hermitian model and their asymptotic behaviour analysed. We then focus on non-Hermitian systems with a real adiabatic parameter and study the LZS interferometry formed out of two identical avoided level crossings. Four distinctive cases of interferometry are identified and the analytic formulae for the transition probabilities are calculated for each case. The differences and similarities between the non-Hermitian case and its Hermitian counterpart are emphasised. In particular, the geometrical phase originated from the sign change of the mass term at the two level crossings is still present in the non-Hermitian system, indicating its robustness against the non-Hermiticity. We further apply our non-Hermitian LZS theory to describing the Bloch oscillation in one-dimensional parity-time $(\mathcal{PT})$ reversal symmetric non-Hermitian Su-Schrieffer-Heeger model and propose an experimental scheme to simulate such dynamics using photonic waveguide arrays. The Landau-Zener transition, as well as the LZS interferometry, can be visualised through the beam intensity profile and the transition probabilitiess measured by the centre of mass of the profile.

quant-ph↗