arXiv subjects

Zejian Li

Publications and source records attributed to Zejian Li.

At least 19 recordsLinked to original sources

arXiv2026

Quantum and Classical Potts Criticality in Driven-Dissipative Bosonic Lattices

The emergence of equilibrium universality from intrinsically nonequilibrium dynamics is a fundamental open problem. Bose-Hubbard lattices realized in photonic and circuit-QED platforms provide a versatile setting to engineer nonlinear interactions, dissipation, and multiphoton processes. Here we investigate a Bose-Hubbard lattice subject to three-photon parametric driving, whose nonequilibrium steady state spontaneously breaks a $\mathbb Z_3$ symmetry and realizes the criticality of the three-state Potts model, a three-state generalization of the Ising model. Using a variational phase-space approach with systematically controllable accuracy based on a Variational Multi-Gaussian ansatz, we perform finite-size scaling analyses in one and two spatial dimensions. We find that, in two-dimensional lattices with single-photon losses, the nonequilibrium steady-state transition belongs to the universality class of the 2D classical three-state Potts model. In contrast, in one-dimensional lattices with three-photon losses, the transition is governed by the one-dimensional quantum three-state Potts universality class. These results establish driven-dissipative bosonic lattices as a platform for emergent Potts criticality and identify multiphoton dissipation as a mechanism that promotes nonequilibrium critical behavior from classical to quantum universality classes.

quant-ph↗

arXiv2026

MoWorld: A Flash World Model

The future of World Models depends not only on scaling model capability, but also on scaling practicality and inference efficiency. High-frame-rate inference enables responsive perception, planning, and control in real-world autonomous systems. To this end, we present MoWorld, a cost-effective yet high-performance Flash World Model with an end-to-end framework spanning data generation, pre-training, distillation, and efficient inference, enabling up to 50 FPS real-time interaction with cinematic visual quality without the need of high-end GPUs. To enable large-scale real-world deployment, MoWorld jointly optimizes model capability and cost throughout the entire development pipeline. Specifically, unlike existing approaches that primarily rely on large-scale video corpora, MoWorld is built upon a scalable 3D-native data engine accumulated from our large-scale 3D vision and generative modeling pipeline, enabling the efficient construction of geometrically consistent training data across diverse real-world and synthetic environments. Based on this foundation, a curriculum cross-frame pre-training strategy for stable and scalable World Model learning, an efficient denoising-step distillation algorithm to reduce diffusion training cost, and a mixed-precision parallel inference framework for low-cost real-time deployment. MoWorld is the first real-time interactive World Model built on the Neural Processing Unit (NPU) and can achieves up to 50 FPS in such the devices, enabling practical and efficient deployment at scale. Comprehensive evaluations demonstrate that MoWorld achieves leading performance; notably, its average inference cost is only 30\%-50\% of that of existing World Models, providing a practical foundation for large-scale real-world applications of World Models. We also demonstrate diverse applications of MoWorld.

cs.CV↗

arXiv2026

Mean Flow Distillation: Robust and Stable Distillation for Flow Matching Models

Flow Matching models have demonstrated strong performance across a wide range of generative tasks. However, their reliance on ODE-based iterative sampling incurs substantial computational overhead in inference, which limits their applicability in real-time scenes. While distillation is a promising solution, existing approaches largely borrow from diffusion-based score matching, often failing to exploit the intrinsic geometric structure of flows and suffering from training instability, high variance, and degraded generation quality. In this paper, we propose Mean Flow Distillation (MFD), a novel distillation framework tailored for flow matching models. We theoretically demonstrate that MFD acts as a temporal low-pass filter, effectively suppressing the high-frequency optimization noise inherent in variational score distillation (VSD) while ensuring global trajectory consistency. We further prove the Mean Flow Matching Theorem, establishing that matching expected average velocities is sufficient for strict distribution alignment. Empirically, on challenging tasks of high-dimensional manifolds including 4D occupancy forecasting and text-to-image generation, MFD achieves state-of-the-art performance, enabling high-fidelity single-step generation.

cs.CV↗

arXiv2026

Quantum jump correlations in long-range dissipative spin systems via cluster and cumulant expansions

We characterize nonequilibrium phases in long-range dissipative spin systems through the statistical properties of quantum jump trajectories. While the average dynamics governed by the Lindblad master equation provides access to steady-state expectation values of order parameters, the quantum trajectory framework reveals features encoded in the spatial and temporal correlations of detection events. Focusing on a model exhibiting a paramagnetic-to-ferromagnetic phase transition, we investigate the full counting statistics of quantum jumps using a tilted Lindbladian approach. We combine this with cluster mean-field and cumulant expansion techniques, which allow us to capture, respectively, the short- and long-range structure of jump correlations. In addition, we study the waiting-time distributions of detection events. We show that quantum jump correlations display clear signatures of the underlying phases and reveal distinct dynamical features across the transition. Our results highlight the potential of trajectory-resolved observables as probes of collective behavior in open quantum many-body systems and provide new insights into the role of long-range interactions in shaping nonequilibrium dynamics.

quant-ph↗

arXiv2026

Generalized stochastic spin-wave theory for open quantum spin systems

We propose a semiclassical framework for solving open quantum dynamics in driven-dissipative spin systems. Our method consists of generalized spin-wave approximations tailored to describing quantum trajectories unravelled from the master equation, and generically applies to regimes beyond the reach of conventional spin-wave theories, including short-range interactions and local quantum jumps, enabling the efficient simulation of large-scale interacting spins. We illustrate the versatility of our framework by studying a variable-range driven-dissipative Ising model on a 2D lattice. When the dissipation acts along the drive axis, we find a continuous phase transition breaking the $\mathbb{Z}_2$ symmetry, and demonstrate that the interaction range, when tuned from fully-connected to nearest-neighbour, profoundly alters the universality class of the criticality. With the dissipation along the interaction axis, we show the emergence of a first-order transition. Demonstrated with both state-diffusion and quantum-jump types of trajectory dynamics, our framework provides a powerful toolbox for the efficient semiclassical description of non-equilibrium dynamics and many-body phases in spin systems.

quant-ph↗

arXiv2026

Variational Dynamics of Open Quantum Spin Systems in Phase Space

We introduce a variational method for simulating the dynamics of interacting open quantum spin systems. The method is based on the spin phase-space representation and variationally targets the Husimi-$Q$ function with an ansatz based on a multi-dimensional mixture of spin-coherent states. Crucially, the mixture coefficients are allowed to take negative values, enabling the faithful capture of quantum correlations beyond semiclassical descriptions. The resulting equations of motion are derived from the Dirac-Frenkel variational principle and can be evaluated efficiently without resorting to Monte Carlo sampling by exploiting the analytical structure of the ansatz. As a first application, we demonstrate that this approach accurately captures both the full quantum dynamics and the non-equilibrium steady states of the transverse-field quantum Ising model, in excellent agreement with exact diagonalization. Furthermore, we show that the method scales efficiently to large two-dimensional lattices, a regime that remains challenging for other techniques.

quant-ph↗

arXiv2026

Beyond Binary Preference: Aligning Diffusion Models to Fine-grained Criteria by Decoupling Attributes

Post-training alignment of diffusion models relies on simplified signals, such as scalar rewards or binary preferences. This limits alignment with complex human expertise, which is hierarchical and fine-grained. To address this, we first construct a hierarchical, fine-grained evaluation criteria with domain experts, which decomposes image quality into multiple positive and negative attributes organized in a tree structure. Building on this, we propose a two-stage alignment framework. First, we inject domain knowledge to an auxiliary diffusion model via Supervised Fine-Tuning. Second, we introduce Complex Preference Optimization (CPO) that extends DPO to align the target diffusion to our non-binary, hierarchical criteria. Specifically, we reformulate the alignment problem to simultaneously maximize the probability of positive attributes while minimizing the probability of negative attributes with the auxiliary diffusion. We instantiate our approach in the domain of painting generation and conduct CPO training with an annotated dataset of painting with fine-grained attributes based on our criteria. Extensive experiments demonstrate that CPO significantly enhances generation quality and alignment with expertise, opening new avenues for fine-grained criteria alignment.

cs.CV↗

arXiv2025

Measurement-induced phase transition in interacting bosons from most likely quantum trajectory

We propose a new theoretical method to describe the monitored dynamics of bosonic many-body systems based on the concept of the most likely trajectory. We show how such trajectory can be identified from the probability distribution of quantum trajectories, i.e. measurement readouts, and how it successfully captures the monitored dynamics beyond the average state. We prove the method to be exact in the case of Gaussian theories and then extend it to the interacting Sine-Gordon model. Although no longer exact in this framework, the method captures the dynamics through a self-consistent time-dependent harmonic approximation and reveals an entanglement phase transition in the steady state from an area-law to a logarithmic-law scaling.

quant-ph↗

arXiv2025

Inversion-DPO: Precise and Efficient Post-Training for Diffusion Models

Recent advancements in diffusion models (DMs) have been propelled by alignment methods that post-train models to better conform to human preferences. However, these approaches typically require computation-intensive training of a base model and a reward model, which not only incurs substantial computational overhead but may also compromise model accuracy and training efficiency. To address these limitations, we propose Inversion-DPO, a novel alignment framework that circumvents reward modeling by reformulating Direct Preference Optimization (DPO) with DDIM inversion for DMs. Our method conducts intractable posterior sampling in Diffusion-DPO with the deterministic inversion from winning and losing samples to noise and thus derive a new post-training paradigm. This paradigm eliminates the need for auxiliary reward models or inaccurate appromixation, significantly enhancing both precision and efficiency of training. We apply Inversion-DPO to a basic task of text-to-image generation and a challenging task of compositional image generation. Extensive experiments show substantial performance improvements achieved by Inversion-DPO compared to existing post-training methods and highlight the ability of the trained generative models to generate high-fidelity compositionally coherent images. For the post-training of compostitional image geneation, we curate a paired dataset consisting of 11,140 images with complex structural annotations and comprehensive scores, designed to enhance the compositional capabilities of generative models. Inversion-DPO explores a new avenue for efficient, high-precision alignment in diffusion models, advancing their applicability to complex realistic generation tasks. Our code is available at https://github.com/MIGHTYEZ/Inversion-DPO

cs.CV↗

arXiv2025

Emergent deterministic entanglement dynamics in monitored infinite-range bosonic systems

We study monitored quantum dynamics of infinite-range interacting bosonic systems in the thermodynamic limit. We show that under semiclassical assumptions, the quantum fluctuations along single monitored trajectories adopt a deterministic limit for both quantum-jump and state-diffusion unravelings, and they can be exactly solved. In particular, the hierarchical structure of the equations of motion explains the coincidence of entanglement criticalities and dissipative phase transitions found in previous finite-size numerical studies. We illustrate the findings on a Bose-Hubbard dimer and a collective spin system.

quant-ph↗

arXiv2025

Diffusion Distillation With Direct Preference Optimization For Efficient 3D LiDAR Scene Completion

The application of diffusion models in 3D LiDAR scene completion is limited due to diffusion's slow sampling speed. Score distillation accelerates diffusion sampling but with performance degradation, while post-training with direct policy optimization (DPO) boosts performance using preference data. This paper proposes Distillation-DPO, a novel diffusion distillation framework for LiDAR scene completion with preference aligment. First, the student model generates paired completion scenes with different initial noises. Second, using LiDAR scene evaluation metrics as preference, we construct winning and losing sample pairs. Such construction is reasonable, since most LiDAR scene metrics are informative but non-differentiable to be optimized directly. Third, Distillation-DPO optimizes the student model by exploiting the difference in score functions between the teacher and student models on the paired completion scenes. Such procedure is repeated until convergence. Extensive experiments demonstrate that, compared to state-of-the-art LiDAR scene completion diffusion models, Distillation-DPO achieves higher-quality scene completion while accelerating the completion speed by more than 5-fold. Our method is the first to explore adopting preference learning in distillation to the best of our knowledge and provide insights into preference-aligned distillation. Our code is public available on https://github.com/happyw1nd/DistillationDPO.

cs.CV↗

arXiv2025

Syllables to Scenes: Literary-Guided Free-Viewpoint 3D Scene Synthesis from Japanese Haiku

In the era of the metaverse, where immersive technologies redefine human experiences, translating abstract literary concepts into navigable 3D environments presents a fundamental challenge in preserving semantic and emotional fidelity. This research introduces HaikuVerse, a novel framework for transforming poetic abstraction into spatial representation, with Japanese Haiku serving as an ideal test case due to its sophisticated encapsulation of profound emotions and imagery within minimal text. While existing text-to-3D methods struggle with nuanced interpretations, we present a literary-guided approach that synergizes traditional poetry analysis with advanced generative technologies. Our framework centers on two key innovations: (1) Hierarchical Literary-Criticism Theory Grounded Parsing (H-LCTGP), which captures both explicit imagery and implicit emotional resonance through structured semantic decomposition, and (2) Progressive Dimensional Synthesis (PDS), a multi-stage pipeline that systematically transforms poetic elements into coherent 3D scenes through sequential diffusion processes, geometric optimization, and real-time enhancement. Extensive experiments demonstrate that HaikuVerse significantly outperforms conventional text-to-3D approaches in both literary fidelity and visual quality, establishing a new paradigm for preserving cultural heritage in immersive digital spaces. Project website at: https://syllables-to-scenes.github.io/

cs.CV↗

arXiv2025

Let Human Sketches Help: Empowering Challenging Image Segmentation Task with Freehand Sketches

Sketches, with their expressive potential, allow humans to convey the essence of an object through even a rough contour. For the first time, we harness this expressive potential to improve segmentation performance in challenging tasks like camouflaged object detection (COD). Our approach introduces an innovative sketch-guided interactive segmentation framework, allowing users to intuitively annotate objects with freehand sketches (drawing a rough contour of the object) instead of the traditional bounding boxes or points used in classic interactive segmentation models like SAM. We demonstrate that sketch input can significantly improve performance in existing iterative segmentation methods, outperforming text or bounding box annotations. Additionally, we introduce key modifications to network architectures and a novel sketch augmentation technique to fully harness the power of sketch input and further boost segmentation accuracy. Remarkably, our model' s output can be directly used to train other neural networks, achieving results comparable to pixel-by-pixel annotations--while reducing annotation time by up to 120 times, which shows great potential in democratizing the annotation process and enabling model training with less reliance on resource-intensive, laborious pixel-level annotations. We also present KOSCamo+, the first freehand sketch dataset for camouflaged object detection. The dataset, code, and the labeling tool will be open sourced.

cs.CV↗

arXiv2024

MindScratch: A Visual Programming Support Tool for Classroom Learning Based on Multimodal Generative AI

Programming has become an essential component of K-12 education and serves as a pathway for developing computational thinking skills. Given the complexity of programming and the advanced skills it requires, previous research has introduced user-friendly tools to support young learners. However, our interviews with six programming educators revealed that current tools often fail to reflect classroom learning objectives, offer flexible, high-quality guidance, and foster student creativity. This highlights the need for more adaptive and reflective tools. Therefore, we introduced MindScratch, a multimodal generative AI (GAI) powered visual programming support tool. MindScratch aims to balance structured classroom activities with free programming creation, supporting students in completing creative programming projects based on teacher-set learning objectives while also providing programming scaffolding. Our user study results indicate that, compared to the baseline, MindScratch more effectively helps students achieve high-quality projects aligned with learning objectives. It also enhances students' computational thinking skills and creative thinking. Overall, we believe that GAI-driven educational tools like MindScratch offer students a focused and engaging learning experience.

cs.HC↗

arXiv2024

LAION-SG: An Enhanced Large-Scale Dataset for Training Complex Image-Text Models with Structural Annotations

Recent advances in text-to-image (T2I) generation have shown remarkable success in producing high-quality images from text. However, existing T2I models show decayed performance in compositional image generation involving multiple objects and intricate relationships. We attribute this problem to limitations in existing datasets of image-text pairs, which lack precise inter-object relationship annotations with prompts only. To address this problem, we construct LAION-SG, a large-scale dataset with high-quality structural annotations of scene graphs (SG), which precisely describe attributes and relationships of multiple objects, effectively representing the semantic structure in complex scenes. Based on LAION-SG, we train a new foundation model SDXL-SG to incorporate structural annotation information into the generation process. Extensive experiments show advanced models trained on our LAION-SG boast significant performance improvements in complex scene generation over models on existing datasets. We also introduce CompSG-Bench, a benchmark that evaluates models on compositional image generation, establishing a new standard for this domain. Our annotations with the associated processing code, the foundation model and the benchmark protocol are publicly available at https://github.com/mengcye/LAION-SG.

cs.CV↗

arXiv2024

Distilling Diffusion Models to Efficient 3D LiDAR Scene Completion

Diffusion models have been applied to 3D LiDAR scene completion due to their strong training stability and high completion quality. However, the slow sampling speed limits the practical application of diffusion-based scene completion models since autonomous vehicles require an efficient perception of surrounding environments. This paper proposes a novel distillation method tailored for 3D Li- DAR scene completion models, dubbed ScoreLiDAR, which achieves efficient yet high-quality scene completion. Score- LiDAR enables the distilled model to sample in significantly fewer steps after distillation. To improve completion quality, we also introduce a novel Structural Loss, which encourages the distilled model to capture the geometric structure of the 3D LiDAR scene. The loss contains a scene-wise term constraining the holistic structure and a point-wise term constraining the key landmark points and their relative configuration. Extensive experiments demonstrate that ScoreLiDAR significantly accelerates the completion time from 30.55 to 5.37 seconds per frame (>5x) on SemanticKITTI and achieves superior performance compared to state-of-the-art 3D LiDAR scene completion models. Our model and code are publicly available on https://github.com/happyw1nd/ScoreLiDAR.

cs.CV↗

arXiv2024

Measurement-induced phase transitions in monitored infinite-range interacting systems

A key challenge in observing measurement-induced phase transitions is the mitigation of the post-selection barrier, which causes the reproducibility of specific sequences of measurement readouts--the trajectory--to be exponentially small in system size. Recent studies suggest that some classes of monitored infinite-range systems alleviate this problem by exhibiting a fast saturation of entanglement, resulting in only a polynomial post-selection overhead. This paper answers whether this feature is inherent in infinite-range systems, due to their underlying semiclassical dynamics. We consider three experimentally relevant monitored models: a Tavis-Cummings model, a Superradiance model, and a Bose-Hubbard dimer, each exhibiting non-trivial monitored dynamics. We unveil the occurrence of entanglement phase transitions in these models, showing how the saturation time is strongly affected by bistability regions, which also prevent the mitigation of the post-selection barrier. Finally, we propose experimental realizations of these models, providing a discussion of post selection from an experimental perspective.

quant-ph↗

arXiv2024

Img2CAD: Conditioned 3D CAD Model Generation from Single Image with Structured Visual Geometry

In this paper, we propose Img2CAD, the first approach to our knowledge that uses 2D image inputs to generate CAD models with editable parameters. Unlike existing AI methods for 3D model generation using text or image inputs often rely on mesh-based representations, which are incompatible with CAD tools and lack editability and fine control, Img2CAD enables seamless integration between AI-based 3D reconstruction and CAD software. We have identified an innovative intermediate representation called Structured Visual Geometry (SVG), characterized by vectorized wireframes extracted from objects. This representation significantly enhances the performance of generating conditioned CAD models. Additionally, we introduce two new datasets to further support research in this area: ABC-mono, the largest known dataset comprising over 200,000 3D CAD models with rendered images, and KOCAD, the first dataset featuring real-world captured objects alongside their ground truth CAD models, supporting further research in conditioned CAD model generation.

cs.CV↗