arXiv subjects

Kun Ding

Publications and source records attributed to Kun Ding.

At least 55 records · Page 3Linked to original sources

arXiv2025

Observation of the Exceptional Skin Effect on a Non-Hermitian Flat band

Flat band and non-Hermitian are both significant conceptions in modern physics. In this study, we delve into the behaviours of flat bands in non-Hermitian systems, focusing on the interplay between the flat band and its dispersive counterparts, investigating the exceptional points (EPs) formed by them together, and the non-Hermitian skin effect (NHSE) on the flat band correspondingly generated, which we name as the exceptional skin effect (ESE). Employing non-Hermitian flat band under chiral/sublattice symmetry, where energy remains highly degenerate, we explore their unique properties. Unlike traditional NHSE which requires the enclosing of a non-zero area in the Bloch complex energy spectrum, the ESE on flat band can be generated with a Bloch complex energy spectrum consisting of one single point, i.e. enclosing no area. By analytically tuning non-Hermitian parameters, changes in the complex energy spectrum and Riemann surfaces are observed, revealing the formation of EPs through the hybridization of flat and dispersive bands while maintaining the dimension of the Hilbert subspace. Additionally, the wave functions of flat band exhibit ESE in specific parameter regions, contrary to the existing frameworks. Experimental validation is conducted using an elastic wave system with actively modulated non-Hermitian parameters, showcasing the impact on flat-band states and confirming the formation of ESE due to flat band-dispersive bands hybridization and correspondingly formed EPs. These results offer novel insights into non-Hermitian physics and present potential directions for further researches and applications in this field.

physics.class-ph↗

arXiv2024

Rethinking Comprehensive Benchmark for Chart Understanding: A Perspective from Scientific Literature

Scientific Literature charts often contain complex visual elements, including multi-plot figures, flowcharts, structural diagrams and etc. Evaluating multimodal models using these authentic and intricate charts provides a more accurate assessment of their understanding abilities. However, existing benchmarks face limitations: a narrow range of chart types, overly simplistic template-based questions and visual elements, and inadequate evaluation methods. These shortcomings lead to inflated performance scores that fail to hold up when models encounter real-world scientific charts. To address these challenges, we introduce a new benchmark, Scientific Chart QA (SCI-CQA), which emphasizes flowcharts as a critical yet often overlooked category. To overcome the limitations of chart variety and simplistic visual elements, we curated a dataset of 202,760 image-text pairs from 15 top-tier computer science conferences papers over the past decade. After rigorous filtering, we refined this to 37,607 high-quality charts with contextual information. SCI-CQA also introduces a novel evaluation framework inspired by human exams, encompassing 5,629 carefully curated questions, both objective and open-ended. Additionally, we propose an efficient annotation pipeline that significantly reduces data annotation costs. Finally, we explore context-based chart understanding, highlighting the crucial role of contextual information in solving previously unanswerable questions.

cs.CL↗

arXiv2024

Unveiling Non-Hermitian Spectral Topology in Hyperbolic Lattices with Non-Abelian Translation Symmetry

The hyperbolic lattice (HBL) has emerged as a compelling platform for exploring matter in non-Euclidean space. Among its notable features, the breakdown of the conventional Bloch theorem stands out, prompting a reexamination of band theory, with the determination of spectra for non-Hermitian systems being a prominent example. Here, we develop an approach to determining the spectra under open boundary conditions (OBCs), one of the foundations in non-Hermitian lattices, from the reciprocal space of HBLs. By introducing supercells to encompass states that are allowed by non-Abelian translational groups, we perform analytic continuation and base on the point gap topology to acquire uniform spectra, the universal OBC spectral range. Applying this method to a single-band nonreciprocal model and a reciprocal non-Abelian semimetal model, we reveal higher-dimensional skin effects and topological phase transitions, respectively, demonstrating the feasibility of our method in predicting spectral topology and investigating non-Hermitian physics in HBLs.

cond-mat.mes-hall↗

arXiv2024

The Magnetic Dislocation in Photonics

The dislocation created in the topological material lays the foundation of many significant findings to control light but requires delicate fabrication of the material. To extend its flexibility and reconfigurability, we propose the magnetic dislocation concept and unveil its properties in a representative model, which effectively combines the topological defect and edge mode at the magnetic domain wall. The results include distinct localization modes and robust light trapping phenomena with the rainbow feature where the eigen-energy of each light-trapping state can be linearly tuned by the magnetic dislocation. The conversion from the trapping state to edge modes can be harnessed by further adiabatically pumping light across an amount of the magnitude of the magnetic dislocation. Our work solves a fundamental problem by introducing magnetic dislocation with new light-manipulation flexibility, which may be implemented in a variety of platforms in photonic, acoustics, and optomechanics with dynamic modulations and synthetic dimensions.

physics.optics↗

arXiv2024

Zero-shot Generalizable Incremental Learning for Vision-Language Object Detection

This paper presents Incremental Vision-Language Object Detection (IVLOD), a novel learning task designed to incrementally adapt pre-trained Vision-Language Object Detection Models (VLODMs) to various specialized domains, while simultaneously preserving their zero-shot generalization capabilities for the generalized domain. To address this new challenge, we present the Zero-interference Reparameterizable Adaptation (ZiRa), a novel method that introduces Zero-interference Loss and reparameterization techniques to tackle IVLOD without incurring additional inference costs or a significant increase in memory usage. Comprehensive experiments on COCO and ODinW-13 datasets demonstrate that ZiRa effectively safeguards the zero-shot generalization ability of VLODMs while continuously adapting to new tasks. Specifically, after training on ODinW-13 datasets, ZiRa exhibits superior performance compared to CL-DETR and iDETR, boosting zero-shot generalizability by substantial 13.91 and 8.74 AP, respectively.Our code is available at https://github.com/JarintotionDin/ZiRaGroundingDINO.

cs.CV↗

arXiv2024

A Survey of Low-shot Vision-Language Model Adaptation via Representer Theorem

The advent of pre-trained vision-language foundation models has revolutionized the field of zero/few-shot (i.e., low-shot) image recognition. The key challenge to address under the condition of limited training data is how to fine-tune pre-trained vision-language models in a parameter-efficient manner. Previously, numerous approaches tackling this challenge have been proposed. Meantime, a few survey papers are also published to summarize these works. However, there still lacks a unified computational framework to integrate existing methods together, identify their nature and support in-depth comparison. As such, this survey paper first proposes a unified computational framework from the perspective of Representer Theorem and then derives many of the existing methods by specializing this framework. Thereafter, a comparative analysis is conducted to uncover the differences and relationships between existing methods. Based on the analyses, some possible variants to improve the existing works are presented. As a demonstration, we extend existing methods by modeling inter-class correlation between representers in reproducing kernel Hilbert space (RKHS), which is implemented by exploiting the closed-form solution of kernel ridge regression. Extensive experiments on 11 datasets are conducted to validate the effectiveness of this method. Toward the end of this paper, we discuss the limitations and provide further research directions.

cs.CV↗

arXiv2024

Calibrated Cache Model for Few-Shot Vision-Language Model Adaptation

Cache-based approaches stand out as both effective and efficient for adapting vision-language models (VLMs). Nonetheless, the existing cache model overlooks three crucial aspects. 1) Pre-trained VLMs are mainly optimized for image-text similarity, neglecting the importance of image-image similarity, leading to a gap between pre-training and adaptation. 2) The current cache model is based on the Nadaraya-Watson (N-W) estimator, which disregards the intricate relationships among training samples while constructing weight function. 3) Under the condition of limited samples, the logits generated by cache model are of high uncertainty, directly using these logits without accounting for the confidence could be problematic. This work presents three calibration modules aimed at addressing the above challenges. Similarity Calibration refines the image-image similarity by using unlabeled images. We add a learnable projection layer with residual connection on top of the pre-trained image encoder of CLIP and optimize the parameters by minimizing self-supervised contrastive loss. Weight Calibration introduces a precision matrix into the weight function to adequately model the relation between training samples, transforming the existing cache model to a Gaussian Process (GP) regressor, which could be more accurate than N-W estimator. Confidence Calibration leverages the predictive variances computed by GP Regression to dynamically re-scale the logits of cache model, ensuring that the cache model's outputs are appropriately adjusted based on their confidence levels. Besides, to reduce the high complexity of GPs, we further propose a group-based learning strategy. Integrating the above designs, we propose both training-free and training-required variants. Extensive experiments on 11 few-shot classification datasets validate that the proposed methods can achieve state-of-the-art performance.

cs.CV↗

arXiv2024

Quantum walks of correlated photons in non-Hermitian photonic lattices

Entanglement entropy characterizes the correlation of multi-particles and unveils the crucial features of open quantum systems. However, the experimental realization of exploring entanglement in non-Hermitian systems remains a challenge. In parallel, quantum walks have offered the possibility of studying the underlying mechanisms of non-Hermitian physics, which includes exceptional points, the non-Hermitian skin effect, and non-Bloch phase transitions. Unfortunately, these studies have only involved and prevailingly focused on the behavior of a single particle. Here, we propose and experimentally realize quantum walks of two indistinguishable photons in engineered non-Hermitian photonic lattices. We have successfully observed the unidirectional behavior of quantum walks in the bulk far from the edges induced by the skin effect. Moreover, we experimentally reveal the suppression of entanglement that is caused by the skin effect in non-Hermitian systems. Our study may facilitate a deep understanding of entanglement in open quantum many-body systems that are far from thermal equilibrium.

quant-ph↗

arXiv2024

Weak Distribution Detectors Lead to Stronger Generalizability of Vision-Language Prompt Tuning

We propose a generalized method for boosting the generalization ability of pre-trained vision-language models (VLMs) while fine-tuning on downstream few-shot tasks. The idea is realized by exploiting out-of-distribution (OOD) detection to predict whether a sample belongs to a base distribution or a novel distribution and then using the score generated by a dedicated competition based scoring function to fuse the zero-shot and few-shot classifier. The fused classifier is dynamic, which will bias towards the zero-shot classifier if a sample is more likely from the distribution pre-trained on, leading to improved base-to-novel generalization ability. Our method is performed only in test stage, which is applicable to boost existing methods without time-consuming re-training. Extensive experiments show that even weak distribution detectors can still improve VLMs' generalization ability. Specifically, with the help of OOD detectors, the harmonic mean of CoOp and ProGrad increase by 2.6 and 1.5 percentage points over 11 recognition datasets in the base-to-novel setting.

cs.CV↗

arXiv2024

Prompt Tuning with Soft Context Sharing for Vision-Language Models

Vision-language models have recently shown great potential on many tasks in computer vision. Meanwhile, prior work demonstrates prompt tuning designed for vision-language models could acquire superior performance on few-shot image recognition compared to linear probe, a strong baseline. In practice, many few-shot tasks are inherently correlated, particularly within specialized domains. However, such information is overlooked previously. Inspired by the fact that modeling task relationship by multi-task learning can usually boost performance, we propose a novel method SoftCPT (Soft Context Sharing for Prompt Tuning) to tune pre-trained vision-language models on multiple target few-shot tasks jointly. Specifically, we design a task-shared meta network to generate prompt context for each task using task name together with a learnable task context as input. The parameters of this meta network as well as the task context are tuned on the joint training set of all tasks. As such, the prompt context of all tasks will be shared in a soft manner. Extensive experiments across four multi-task few-shot datasets covering 44 tasks and 1593 categories demonstrate that SoftCPT significantly outperforms single-task prompt tuning methods, highlighting the effectiveness of multi-task learning for vision-language prompt tuning. Code is available at https://github.com/kding1225/softcpt.

cs.CV↗

arXiv2024

Compositional Kronecker Context Optimization for Vision-Language Models

Context Optimization (CoOp) has emerged as a simple yet effective technique for adapting CLIP-like vision-language models to downstream image recognition tasks. Nevertheless, learning compact context with satisfactory base-to-new, domain and cross-task generalization ability while adapting to new tasks is still a challenge. To tackle such a challenge, we propose a lightweight yet generalizable approach termed Compositional Kronecker Context Optimization (CK-CoOp). Technically, the prompt's context words in CK-CoOp are learnable vectors, which are crafted by linearly combining base vectors sourced from a dictionary. These base vectors consist of a non-learnable component obtained by quantizing the weights in the token embedding layer, and a learnable component constructed by applying Kronecker product on several learnable tiny matrices. Intuitively, the compositional structure mitigates the risk of overfitting on training data by remembering more pre-trained knowledge. Meantime, the Kronecker product breaks the non-learnable restrictions of the dictionary, thereby enhancing representation ability with minimal additional parameters. Extensive experiments confirm that CK-CoOp achieves state-of-the-art performance under base-to-new, domain and cross-task generalization evaluation, but also has the metrics of fewer learnable parameters and efficient training and inference speed.

cs.CV↗

arXiv2024

Non-Bloch Theory for Spatiotemporal Photonic Crystals Assisted by Continuum Effective Medium

As one indispensable type of nonreciprocal mechanism, a system with temporal modulations is intrinsically open in the physical sense and inevitably non-Hermitian, but the space and time degrees of freedom are nonseparable in a large variety of circumstances, which restrains the non-Bloch band theory to apply. Here, we investigate the spatially photonic crystals (PhCs) composed of spatiotemporal modulation materials (STMs) and homogeneous media, dubbed as the STM-PhC, wherein the spatial and temporal modulations are deliberately designed to be correlated. To bypass the difficulty of the spatiotemporal correlation, we first employ the effective medium theory to account for the dispersion of fundamental bands under the influence of Floquet sidebands. Based on the dynamical degeneracy splitting viewpoint and continuum generalized Brillouin zone condition, we then analytically give the criteria for the existence of the non-Hermitian skin effect in the STM. Assisted by developing a numerical method that embeds the plane wave expansion in the transfer matrix, we establish the non-Bloch band theory for the low-frequency Floquet bands in the STM-PhCs, in which the central is the identification of the generalized Brillouin zone. We finally delve into the topological properties, including non-Bloch Zak phases and delocalization of topologically edge states. Our work validates that effective medium assists the non-Bloch band theory applied to the STM-PhCs, which delivers a prescription to broaden the horizons of non-Bloch theory.

physics.optics↗

arXiv2023

PAD: Self-Supervised Pre-Training with Patchwise-Scale Adapter for Infrared Images

Self-supervised learning (SSL) for RGB images has achieved significant success, yet there is still limited research on SSL for infrared images, primarily due to three prominent challenges: 1) the lack of a suitable large-scale infrared pre-training dataset, 2) the distinctiveness of non-iconic infrared images rendering common pre-training tasks like masked image modeling (MIM) less effective, and 3) the scarcity of fine-grained textures making it particularly challenging to learn general image features. To address these issues, we construct a Multi-Scene Infrared Pre-training (MSIP) dataset comprising 178,756 images, and introduce object-sensitive random RoI cropping, an image preprocessing method, to tackle the challenge posed by non-iconic images. To alleviate the impact of weak textures on feature learning, we propose a pre-training paradigm called Pre-training with ADapter (PAD), which uses adapters to learn domain-specific features while freezing parameters pre-trained on ImageNet to retain the general feature extraction capability. This new paradigm is applicable to any transformer-based SSL method. Furthermore, to achieve more flexible coordination between pre-trained and newly-learned features in different layers and patches, a patchwise-scale adapter with dynamically learnable scale factors is introduced. Extensive experiments on three downstream tasks show that PAD, with only 1.23M pre-trainable parameters, outperforms other baseline paradigms including continual full pre-training on MSIP. Our code and dataset are available at https://github.com/casiatao/PAD.

cs.CV↗

arXiv2023

Superior ferroelectricity and nonlinear optical response in a hybrid germanium iodide hexagonal perovskite

Abundant chemical diversity and structural tunability make organic-inorganic hybrid perovskites (OIHPs) a rich ore for ferroelectrics. However, compared with their inorganic counterparts such as BaTiO$_3$, their ferroelectric key properties, including large spontaneous polarization ($P_s$), low coercive field ($E_c$), and strong second harmonic generation (SHG) response, have long been great challenges, which hinder their commercial applications. Here, a quasi-one-dimensional OIHP DMAGeI$_3$ (DMA=Dimethylamine) is reported, with notable ferroelectric attributes at room temperature: a large $P_s$ of 24.14 $μ$C/cm$^2$ (on a par with BaTiO$_3$), a low $E_c$ below 2.2 kV/cm, and the strongest SHG intensity in OIHP family (about 12 times of KH$_2$PO$_4$ (KDP)). Revealed by the first-principles calculations, its large $P_s$ originates from the synergistic effects of the stereochemically active $4s^2$ lone pair of Ge$^{2+}$ and the ordering of organic cations, and its low kinetic energy barrier of small DMA cations results in a low $E_c$. Our work brings the comprehensive ferroelectric performances of OIHPs to a comparable level with commercial inorganic ferroelectric perovskites.

cond-mat.mtrl-sci↗

arXiv2023

Exceptional Non-Abelian Topology in Multiband Non-Hermitian Systems

Defective spectral degeneracy, known as exceptional point (EP), lies at the heart of various intriguing phenomena in optics, acoustics, and other nonconservative systems. Despite extensive studies in the past two decades, the \textit{collective} behaviors (e.g., annihilation, coalescence, braiding, etc.) involving multiple exceptional points or lines and their interplay have been rarely understood. Here we put forward a universal non-Abelian conservation rule governing these collective behaviors in generic multiband non-Hermitian systems and uncover several counterintuitive phenomena. We demonstrate that two EPs with opposite charges (even the pairwise created) do not necessarily annihilate, depending on how they approach each other. Furthermore, we unveil that the conservation rule imposes strict constraints on the permissible exceptional-line configurations. It excludes structures like Hopf link yet permits novel staggered rings composed of noncommutative exceptional lines. These intriguing phenomena are illustrated by concrete models which could be readily implemented in platforms like coupled acoustic cavities, optical waveguides, and ring resonators. Our findings lay the cornerstone for a comprehensive understanding of the exceptional non-Abelian topology and shed light on the versatile manipulations and applications based on exceptional degeneracies in nonconservative systems.

cond-mat.mes-hall↗

arXiv2023

Noncontact frictional force between surfaces by peristaltic permittivity modulation

In this study, we reveal noncontact frictional forces between surfaces in the presence of peristaltic permittivity modulation. Our setup comprises a conducting medium, an air gap, and a dielectric substrate on which we have a space-time-modulated grating that emits electromagnetic radiation. The radiation receives energy and momentum from the grating, which is eventually absorbed by the conducting medium or propagates away from the grating on the dielectric side, resulting in electromagnetic power loss and lateral forces at the surfaces.

physics.optics↗

arXiv2023

Experimental Realization of Geometry-Dependent Skin Effect in a Reciprocal Two-Dimensional Lattice

Recent studies of non-Hermitian periodic lattices unveiled the non-Hermitian skin effect (NHSE), in which the bulk modes under the periodic boundary conditions (PBC) become skin modes under open boundary conditions (OBC). The NHSE is a topological effect owing to the non-trivial spectral winding, and such spectral behaviors appear naturally in non-reciprocal systems. Hence prevailing approaches rely on non-reciprocity to achieve the NHSE. Here, we report the experimental realization of the geometry-dependent skin effect (GDSE) in a two-dimensional (2D) reciprocal system, in which the skin effect occurs only at boundaries whose macroscopic symmetry mismatches with the lattice symmetry. The role of spectral reciprocity and symmetry is revealed by connecting reflective channels at given boundaries with the spectral topology of the PBC spectrum. Our work highlights the vital role of reciprocity and macroscopic geometry on the NHSE in systems with dimensions larger than one and opens new routes for wave structuring using non-Hermitian effects.

cond-mat.other↗

arXiv2022

Realization and Topological Properties of Third-Order Exceptional Lines Embedded in Exceptional Surfaces

As the counterpart of Hermitian nodal structures, the geometry formed by exceptional points (EPs), such as exceptional lines (ELs), entails intriguing spectral topology. We report the experimental realization of order-3 exceptional lines (EL3) that are entirely embedded in order-2 exceptional surfaces (ES2) in a three-dimensional periodic synthetic momentum space. The EL3 and the concomitant ES2, together with the topology of the underlying space, prohibit the evaluation of their topology in the eigenvalue manifold by prevailing topological characterization methods. We resolve this issue by defining a winding number that associates with the resultants of the Hamiltonian. This resultant winding number detects EL3 but ignores the ES2, allowing the diagnosis of the topological currents carried by the EL3, which enables the prediction of their evolution under perturbations. Our results exemplify unprecedented topology of higher-order exceptional geometries and may inspire new non-Hermitian topological applications.

cond-mat.mes-hall↗