SearcharxivSearch

arXiv subjects

Miaomiao Zhao

Publications and source records attributed to Miaomiao Zhao.

7 recordsLinked to original sources

GeoDiff3D: Self-Supervised 3D Scene Generation with Geometry-Constrained 2D Diffusion Guidance

3D scene generation is a core technology for gaming, film/VFX, and VR/AR. Growing demand for rapid iteration, high-fidelity detail, and accessible content creation has further increased interest in this area. Existing methods broadly follow two paradigms - indirect 2D-to-3D reconstruction and direct 3D generation - but both are limited by weak structural modeling and heavy reliance on large-scale ground-truth supervision, often producing structural artifacts, geometric inconsistencies, and degraded high-frequency details in complex scenes. We propose GeoDiff3D, an efficient self-supervised framework that uses coarse geometry as a structural anchor and a geometry-constrained 2D diffusion model to provide texture-rich reference images. Importantly, GeoDiff3D does not require strict multi-view consistency of the diffusion-generated references and remains robust to the resulting noisy, inconsistent guidance. We further introduce voxel-aligned 3D feature aggregation and dual self-supervision to maintain scene coherence and fine details while substantially reducing dependence on labeled data. GeoDiff3D also trains with low computational cost and enables fast, high-quality 3D scene generation. Extensive experiments on challenging scenes show improved generalization and generation quality over existing baselines, offering a practical solution for accessible and efficient 3D scene construction.

cs.CV

Altermagnetism in exactly solvable model: the Ising-Kondo lattice model

Altermagnet (AM), a recently identified class of collinear magnet, has garnered significant attention due to its unique combination of zero net magnetization and spin-split energy bands, leading to a variety of novel physical phenomena. Using numerically exact lattice Monte Carlo simulations, we investigate AM-like phases within the Ising-Kondo lattice model which is commonly employed to describe heavy-fermion materials. By incorporating an alternating next-nearest-neighbor hopping (NNNH) term, which arises from the influence of non-magnetic atoms in altermagnetic candidate materials, our results reveal key signatures of AM-like states, including spin-splitting quasiparticle bands and spectral functions, and demonstrate that d-wave AM remains stable across a broad range of interaction strengths, doping levels, NNNH amplitudes and temperatures, highlighting its robustness. Furthermore, through an analysis of non-magnetic impurity effects, we further confirm the d-wave symmetry of the AM phase. These findings establish a solid theoretical foundation for exploring AM-like phases in f-electron compounds, paving the way for future investigations into their exotic magnetic and electronic properties.

cond-mat.str-el

MDiff4STR: Mask Diffusion Model for Scene Text Recognition

Mask Diffusion Models (MDMs) have recently emerged as a promising alternative to auto-regressive models (ARMs) for vision-language tasks, owing to their flexible balance of efficiency and accuracy. In this paper, for the first time, we introduce MDMs into the Scene Text Recognition (STR) task. We show that vanilla MDM lags behind ARMs in terms of accuracy, although it improves recognition efficiency. To bridge this gap, we propose MDiff4STR, a Mask Diffusion model enhanced with two key improvement strategies tailored for STR. Specifically, we identify two key challenges in applying MDMs to STR: noising gap between training and inference, and overconfident predictions during inference. Both significantly hinder the performance of MDMs. To mitigate the first issue, we develop six noising strategies that better align training with inference behavior. For the second, we propose a token-replacement noise mechanism that provides a non-mask noise type, encouraging the model to reconsider and revise overly confident but incorrect predictions. We conduct extensive evaluations of MDiff4STR on both standard and challenging STR benchmarks, covering diverse scenarios including irregular, artistic, occluded, and Chinese text, as well as whether the use of pretraining. Across these settings, MDiff4STR consistently outperforms popular STR models, surpassing state-of-the-art ARMs in accuracy, while maintaining fast inference with only three denoising steps. Code: https://github.com/Topdu/OpenOCR.

cs.CV

Altermagnetism in Heavy Fermion Systems: Mean-Field study on Kondo Lattice

Recently, a novel collinear magnet, i.e. the altermagnet (AM), with spin-splitting energy band and zero net magnetization have attracted great interest due to its potential spintronic applications. Here, we demonstrate AM-like phases in a microscopic Kondo lattice (KL) model with an alternating next-nearest-neighbor-hopping (NNNH). Such alternating NNNH take nonmagnetic atoms, neglected in usual antiferromagnetism study, into account when encountering real-life candidate AM materials. With the framework of fermionic parton mean-field theory, we find three different ground-states for the half-filling KL: 1) a $d$-wave AM state; 2) a coexistent phase with both $d$-wave AM and intrinsic Kondo screening effect; 3) a Kondo insulator. The AM-like states are characterized by their spin-splitting quasiparticle bands, Fermi surface, spin-resolved distribution function and conductivity. It is suggested that the magnetic quantum oscillation, scanning tunneling microscopy and charge transport measurement can detect those AM-like phases. We hope the present work may be useful for exploring AM-like phases in $f$-electron compounds such as CeNiAsO and Ce$_{4}$X$_{3}$(X=As,Sb,Bi).

cond-mat.str-el

Topic Review: Hatsugai-Kohmoto models: Exactly solvable playground for Mottness and Non-Fermi Liquid

This pedagogic review aims to give a gentle introduction to an exactly solvable model, the Hatsugai-Kohmoto (HK) model, which has infinite-ranged interaction but conserves the center of mass. Although this model is invented in 1992, intensive studies on its properties ranging from unconventional superconductivity, topological ordered states to non-Fermi liquid behaviors are made since 2020. We focus on its emergent non-Fermi liquid behavior and provide discussion on its thermodynamics, single-particle and two-particle correlation functions. Perturbation around solvable limit has also been explored with the help of perturbation theory, renormalization group and exact diagonalization calculation. We hope the present review will be helpful for graduate students or researchers interested in HK-like models or more generic strongly correlated electron systems.

cond-mat.str-el

Friedel oscillation in non-Fermi liquid: Lesson from exactly solvable Hatsugai-Kohmoto model

When non-magnetic impurity immerses in Fermi sea, a regular modulation of charge density around impurity will appear and such phenomena is called Friedel oscillation (FO). Although both Luttinger liquid and Landau Fermi liquid show such characteristic oscillation, FO in generic non-Fermi liquid (NFL) phase is still largely unknown. Here, we show that FO indeed exists in NFL state of an exactly solvable model, i.e. the Hatsugai-Kohmoto model which has been intensively explored in recent years. Combining T-matrix approximation and linear-response-theory, an interesting picture emerges, if two interaction-induced quasi-particles bands in NFL are partially occupied, FO in this situation is determined by a novel structure in momentum space, i.e. the 'average Fermi surface' (average over two quasi-particle Fermi surface), which highlights the inter-band particle-hole excitation. We hope our study here provides a counterintuitive example in which FO with Fermi surface coexists with NFL quasi-particle, and it may be useful to detect hidden 'average Fermi surface' structure in other correlated electron systems.

cond-mat.str-el

The process of superradiant phase transition for quantum Rabi model in view of nonclassical states

The ground state of quantum Rabi model (QRM) exhibits rich nonclassical states including squeezed state, cat state, and entangled state in different parameter regimes. In this paper, we firstly use the polaron picture to figure out the process of superradiant phase transition (SPT) in view of the nonclassical states. According to the polaron wave functions, the squeezed state extends to whole parameter regimes, and a small but non-zero weighted antipolaron is necessary to form novel semi-cat states. Moreover, the behavior of superradiance in the QRM can be viewed as the same displacement of the cat states from zero to a finite value, while the ground state becomes entangled state resulting from the emergence of spin-up state. On the other hand, the nonclassical states can be intuitively characterized by the Wigner functions with analytical expressions in the polaron picture, and the Wigner negativity is also evaluated to measure the nonclassicality. Based on the squeezing and superradiance, a classification of the coupling strength for the nonclassical states containing in the ground state is presented, and the process of the SPT is also revealed by the photon number distribution in Fock space, which is important to understand the fundamental quantum physics in the QRM. Our work provides a guidance for preparing the nonclassical states in experiments based on the QRM.

quant-ph