arXiv subjects

Kun Huang

Publications and source records attributed to Kun Huang.

At least 73 records · Page 4Linked to original sources

arXiv2026

Wide-field mid-infrared edge-enhanced upconversion imaging

Edge-enhanced imaging is critical for visualizing weakly absorbing and transparent objects. Extending this functionality into the mid-infrared (MIR) region enables chemical sensitivity and improved imaging performance for biomedical, material, and remote-sensing applications. Here, we present a wide-field MIR edge-enhanced upconversion imaging system that integrates vortex-pump complex-amplitude engineering with aperiodic quasi-phase matching. In contrast to the bright-field modality, the wide-field edge-enhanced operation shows sensitive dependence on the crystal position relative to the Fourier plane. The system achieves single-shot operation with a 25-mm field of view and 79-$μ$m spatial resolution, yielding a record-high space-bandwidth product of $7.9 \times 10^4$. We show that this capability enables direct visualization of phase gradients in transparent optical elements and enhances structural contrast in biological specimens. The demonstrated architecture combines high sensitivity, spectral specificity, and robust edge detection, offering a promising route toward advanced MIR imaging in industrial inspection and biomedical diagnostics.

physics.optics↗

arXiv2026

High-order mid-infrared nonlinear topological differentiator

High-order edge-enhanced imaging enables precise feature localization and effective background suppression, offering a powerful tool for real-time recognition and high-contrast visualization. Extending this capability to the mid-infrared (MIR) regime is particularly valuable for applications such as biomedical diagnostics, material inspection, and remote sensing, yet remains limited by inadequate spatial-frequency modulation fidelity and low detection sensitivity. Here, we demonstrate a high-sensitivity MIR upconversion differentiator operating at 3 $μ$m, which achieves isotropic high-order edge enhancement by optically imprinting topological complex-amplitude patterns onto MIR Fourier components via nonlinear parametric interaction. Vortex transfer functions $t(k_r, ϕ) \propto k_r^\ell e^{i\ellϕ}$ are precisely encoded on a phase-only spatial light modulator to enable tunable MIR differentiation from first- to fourth- order, with real-time switching at up to 60 Hz. Benefiting from a low-noise upconversion process and a single-photon-sensitive silicon camera, the system achieves high-contrast edge imaging under low-light conditions. Experimental results confirm accurate edge extraction and background suppression for both amplitude and phase objects, hence underscoring its potential for noninvasive diagnostics and label-free material analysis.

physics.optics↗

arXiv2026

The consecutive lifting-projection flow as an approximation of Boltzmann and Landau flow

We introduce the consecutive lifting-projection (LP) flow as a novel approximation framework for the spatially homogeneous Boltzmann and Landau equations. The key idea is to lift the nonlinear collision operator to a higher dimensional linear Kac master equation on spheres, evolve this lifted equation in time, and project the solution back to the lower dimensional velocity space. The resulting LP flow is a tangent flow to the original kinetic dynamics and admits a clear semigroup structure. We show that the consecutive LP flow preserves mass, momentum, and energy, satisfies an entropy dissipation property, and converges to the correct Maxwellian equilibrium. In addition, the lifting removes the intrinsic nonlinearity of the collision operator and enables explicit analytical representations of the solution. For Maxwell molecules, we provide an error estimate quantifying the accuracy over finite time intervals. The framework provides a concise and general methodology for constructing reliable numerical solvers in kinetic theory. It unifies existing explicit discretizations, which helps understanding numerical stability and clarifying the trade-off between conservation and positivity. More importantly, it enables the development of new schemes. In particular, we propose the Green's function method, which is not only unconditionally stable, but also perfectly compatible with fast spectral discretizations.

math.NA↗

arXiv2026

Distributed Normal Map-based Stochastic Proximal Gradient Methods over Networks

Consider $n$ agents connected over a network collaborating to minimize the average of their local cost functions combined with a common nonsmooth function. This paper introduces a unified algorithmic framework for solving such a problem through distributed stochastic proximal gradient methods, leveraging the normal map update scheme. Within this framework, we propose two new algorithms, termed Normal Map-based Distributed Stochastic Gradient Tracking (norM-DSGT) and Normal Map-based Exact Diffusion (norM-ED). We demonstrate that both methods can asymptotically achieve comparable convergence rates to the centralized stochastic proximal gradient descent method under a general variance condition on the stochastic gradients. Additionally, the number of iterations required for norM-ED to achieve such a rate (i.e., the transient time) behaves as $\mathcal{O}(n^{3}/(1-λ)^2)$ for minimizing composite objective functions, matching the performance of the non-proximal ED algorithm. Here $1-λ$ denotes the spectral gap of the mixing matrix related to the underlying network topology. To our knowledge, such a convergence result is state-of-the-art for the considered composite problem. Under the same condition, norM-DSGT enjoys a transient time of $\mathcal{O}(\max\{n^3/(1-λ)^2, n/(1-λ)^3\})$, which matches that of the non-proximal DSGT algorithm and norM-ED under the condition $(1-λ)^{-1}=\mathcal{O}(n^{2})$.

math.OC↗

arXiv2026

Spatiotemporal flat optics for terabit-per-second single-channel data transmission

Exponential growth in global data traffic demands ever-increasing transmission rates--a pursuit fundamentally constrained by the physical limitations of digital-to-analog converters (DACs). Existing strategies to overcome this bottleneck, such as multi-DAC arrays and optical time-division multiplexing, inevitably introduce system complexity and coordination overhead. Here we demonstrate an all-optical spatiotemporal transmitter that generates controllable high-repetition information-carrying femtosecond pulses at the focus of a phase-modulated planar diffractive lens (PDL) through optical-path-induced spatial-to-temporal conversion. Each pulse serves as an information bit, encoding binary data via on-axis focal intensity states corresponding to '0' and '1', achieved by switching between topological and constant phase modulations. High experimental orthogonality between arbitrary bits enables nearly error-free transmission of 15X15-pixel grayscale (8-bit coding) and colour (9-bit coding) images at a record-high single-channel rate of approximately 3 terabits per second (Tbit/s). Free from electronic and coordination bottlenecks, this all-optical transmitter establishes a scalable high-speed single-channel pathway toward ultrahigh-capacity optical communication.

physics.optics↗

arXiv2026

MobileIPL: Enhancing Mobile Agents Thinking Process via Iterative Preference Learning

The Chain of Action-Planning Thoughts (CoaT) paradigm has been shown to improve the reasoning performance of VLM-based mobile agents in GUI tasks. However, the scarcity of diverse CoaT trajectories limits the expressiveness and generalization ability of such agents. While self-training is commonly employed to address data scarcity, existing approaches either overlook the correctness of intermediate reasoning steps or depend on expensive process-level annotations to construct process reward models (PRM). To address the above problems, we propose an Iterative Preference Learning (IPL) that constructs a CoaT-tree through interative sampling, scores leaf nodes using rule-based reward, and backpropagates feedback to derive Thinking-level Direct Preference Optimization (T-DPO) pairs. To prevent overfitting during warm-up supervised fine-tuning, we further introduce a three-stage instruction evolution, which leverages GPT-4o to generate diverse Q\&A pairs based on real mobile UI screenshots, enhancing both generality and layout understanding. Experiments on three standard Mobile GUI-agent benchmarks demonstrate that our agent MobileIPL outperforms strong baselines, including continual pretraining models such as OS-ATLAS and UI-TARS. It achieves state-of-the-art performance across three standard Mobile GUI-Agents benchmarks and shows strong generalization to out-of-domain scenarios.

cs.CL↗

arXiv2026

CoME: Empowering Channel-of-Mobile-Experts with Informative Hybrid-Capabilities Reasoning

Mobile Agents can autonomously execute user instructions, which requires hybrid-capabilities reasoning, including screen summary, subtask planning, action decision and action function. However, existing agents struggle to achieve both decoupled enhancement and balanced integration of these capabilities. To address these challenges, we propose Channel-of-Mobile-Experts (CoME), a novel agent architecture consisting of four distinct experts, each aligned with a specific reasoning stage, CoME activates the corresponding expert to generate output tokens in each reasoning stage via output-oriented activation. To empower CoME with hybrid-capabilities reasoning, we introduce a progressive training strategy: Expert-FT enables decoupling and enhancement of different experts' capability; Router-FT aligns expert activation with the different reasoning stage; CoT-FT facilitates seamless collaboration and balanced optimization across multiple capabilities. To mitigate error propagation in hybrid-capabilities reasoning, we propose InfoGain-Driven DPO (Info-DPO), which uses information gain to evaluate the contribution of each intermediate step, thereby guiding CoME toward more informative reasoning. Comprehensive experiments show that CoME outperforms dense mobile agents and MoE methods on both AITZ and AMEX datasets.

cs.CL↗

arXiv2026

Broadband terahertz comb with sub-Hz comb linewidth

Terahertz (THz) frequency combs are increasingly essential for spectroscopy, metrology, and quantum science. However, generating a dense array of evenly spaced ultra-narrow THz comb lines is challenging. Here, we demonstrate broadband THz comb generation using a photoconductive antenna that transfers a noise-suppressed near-infrared electro-optical (EO) comb into the THz domain. Our noise-suppression strategy, leveraging soliton self-frequency shift and spectral filtering, effectively suppresses EO comb phase noise without requiring active stabilization. The resulting THz comb exhibits broad spectral coverage (0.05-4 THz), narrow comb linewidths (0.3 Hz at the Fourier-transform limit), and excellent frequency stability (8.6*10^-14 at 1-second integration). We further demonstrate asynchronous THz time-domain spectroscopy, resolving ~36,000 comb lines with 50 MHz spacing. Crucially, the inherent frequency agility of the EO comb enables rapid and wide-range tuning of the THz comb line spacing. These attributes position our THz comb as a versatile tool for high-resolution molecular spectroscopy and precision THz metrology.

physics.optics↗

arXiv2026

Cross360: 360° Monocular Depth Estimation via Cross Projections Across Scales

360° depth estimation is a challenging research problem due to the difficulty of finding a representation that both preserves global continuity and avoids distortion in spherical images. Existing methods attempt to leverage complementary information from multiple projections, but struggle with balancing global and local consistency. Their local patch features have limited global perception, and the combined global representation does not address discrepancies in feature extraction at the boundaries between patches. To address these issues, we propose Cross360, a novel cross-attention-based architecture integrating local and global information using less-distorted tangent patches along with equirectangular features. Our Cross Projection Feature Alignment module employs cross-attention to align local tangent projection features with the equirectangular projection's 360° field of view, ensuring each tangent projection patch is aware of the global context. Additionally, our Progressive Feature Aggregation with Attention module refines multi-scaled features progressively, enhancing depth estimation accuracy. Cross360 significantly outperforms existing methods across most benchmark datasets, especially those in which the entire 360° image is available, demonstrating its effectiveness in accurate and globally consistent depth estimation. The code and model are available at https://github.com/huangkun101230/Cross360.

cs.CV↗

arXiv2026

PanoNormal: Monocular Indoor 360° Surface Normal Estimation

The presence of spherical distortion in equirectangular projection (ERP) images presents a persistent challenge in dense regression tasks such as surface normal estimation. Although it may appear straightforward to repurpose architectures developed for 360° depth estimation, our empirical findings indicate that such models yield suboptimal performance when applied to surface normal prediction. This is largely attributed to their architectural bias toward capturing global scene layout, which comes at the expense of the fine-grained local geometric cues that are critical for accurate surface orientation estimation. While convolutional neural networks (CNNs) have been employed to mitigate spherical distortion, their fixed receptive fields limit their ability to capture holistic scene structure. Conversely, vision transformers (ViTs) are capable of modeling long-range dependencies via global self-attention, but often fail to preserve high-frequency local detail. To address these limitations, we propose \textit{PanoNormal}, a monocular surface normal estimation architecture for 360° images that integrates the complementary strengths of CNNs and ViTs. In particular, we design a multi-level global self-attention mechanism that explicitly accounts for the spherical feature distribution, enabling our model to recover both global contextual structure and local geometric details. Experimental results demonstrate that our method not only achieves state-of-the-art performance on several benchmark 360° datasets, but also significantly outperforms adapted depth estimation models on the task of surface normal prediction. The code and model are available at https://github.com/huangkun101230/PanoNormal.

cs.CV↗

arXiv2025

A dispersion-driven 3D color near-eye meta-display

Chromatic dispersion, an inherent wavelength-dependent phenomenon in optical systems, has traditionally been regarded as a detrimental effect to be minimized in imaging and display. Here, we present a paradigm shift by deliberately engineering and harnessing metalens dispersion as a functional mechanism for three-dimensional (3D) near-eye displays. Specifically, we exploit lateral dispersion to transform transverse offset between green and red objects into image-space angular separations that make their images intersected virtually, thereby creating color-merged 3D virtual-image perception. This meta-display architecture preserves compactness of conventional planar display while exhibiting less data requirements and lower hardware complexity than other near-eye 3D displays. Experimentally, we demonstrate a multi-color near-eye 3D system achieving an 11° field of view, 22 pixels-per-degree angular resolution, 0.9 m depth of field, and 19 distinct image planes. This work establishes a new pathway for metasurfaces toward visual displays and highlights great potential for future virtual/augmented reality.

physics.optics↗

arXiv2025

A Joint Auction Framework with Externalities and Adaptation

Recently, joint advertising has gained significant attention as an effective approach to enhancing the efficiency and revenue of advertising slot allocation. Unlike traditional advertising, which allocates advertising slots exclusively to a single advertiser, joint advertising displays advertisements from brands and stores that have established a joint selling relationship within the same advertising slot. However, existing approaches often struggle to accommodate both joint and traditional advertising frameworks, thereby limiting the revenue potential and generalizability of joint advertising. Furthermore, these methods are constrained by two critical limitations: they generally neglect the influence of global externalities, and they fail to address the bidding variability stemming from multi-party advertiser participation. Collectively, these limitations present substantial challenges to the design of joint auction mechanisms. To address these challenges, we propose a Joint Auction Framework incorporating Externalities and Adaptation, and leverage the automated mechanism design (AMD) method through our proposed JEANet to compute joint auction mechanisms that satisfy the conditions of individual rationality (IR) and approximate dominant strategy incentive compatibility (DSIC). As the first AMD method to integrate global externalities into joint auctions, JEANet dynamically adapts to the bidding characteristics of multi-party advertiser and enables unified auctions that integrate both joint and traditional advertising. Extensive experimental results demonstrate that JEANet outperforms state-of-the-art baselines in multi-slot joint auctions.

cs.GT↗

arXiv2025

An all-optical convolutional neural network for image identification

In modern artificial intelligence, convolutional neural networks (CNNs) have become a cornerstone for visual and perceptual tasks. However, their implementation on conventional electronic hardware faces fundamental bottlenecks in speed and energy efficiency due to resistive and capacitive losses. Photonic alternatives offer a promising route, yet the difficulty of realizing optical nonlinearities has prevented the realization of all-optical CNNs capable of end-to-end image classification. Here, we demonstrate an all-optical CNN that bypasses the need for explicit optical nonlinear activations. Our architecture comprises a single spatial-differentiation convolutional stage--using 24 directional kernels spanning 360°, along with a mean-filtering kernel--followed by a diffractive fully-connected layer. The directional convolution enhances feature selectivity, suppresses noise and crosstalk, and simplifies the classification task, allowing the weak nonlinearity inherent in optical diffraction to achieve high accuracy. We report experimentally classification accuracies of 86.8% on handwritten digits (MNIST) and 94.8% on a ten-class gesture dataset. The system delivers a computational throughput of 1.13X10^5 tera-operations per second (TOPS) and an energy efficiency of 1.51X10^3 TOPS/W--the highest reported among CNN hardware--with the potential to improve by a further 5-6 orders of magnitude using nanosecond-scale detectors. This work establishes a scalable pathway toward ultralow-latency, ultralow-energy vision processing for real-time intelligent systems.

physics.optics↗

arXiv2025

A 160 ° x 160 ° Dynamic Holographic Meta-Projector

Holography can reconstruct immersive light fields for virtual and augmented reality by modulating optical wavefront. Due to huge pixel sizes, current spatial light modulators (SLMs) have small field-of-view (FOV) for holographic displays. Despite various methods for etendue expansion, the largest full-screen FOV for dynamic holography is only 70 ° X 70 °, which remains insufficient for large-scale, high-resolution, three-dimensional displays. Here, we report a pixel-interpolation-assisted holographic meta-projector that substantially expands the FOV by integrating multiple subwavelength metasurface pixels within each microscale pixel of a traditional SLM. Leveraging large-angle diffraction of the metasurface and implementing k-space distortion correction for ultra-wide angles, we experimentally demonstrate dynamic holographic image reconstruction with a FOV of 160 ° X 160 ° -equivalent to a system numerical aperture of 0.985-at a high framerate of 60 Hz, surpassing the temporal resolution threshold of human vision. This system represents the state-of-the-art near-full-screen holographic dynamic display, thereby opening the door to high-dynamic-range and large-FOV holographic displays.

physics.optics↗

arXiv2025

DCMIL: A Progressive Representation Learning of Whole Slide Images for Cancer Prognosis Analysis

The burgeoning discipline of computational pathology shows promise in harnessing whole slide images (WSIs) to quantify morphological heterogeneity and develop objective prognostic modes for human cancers. However, progress is impeded by the computational bottleneck of gigapixel-size inputs and the scarcity of dense manual annotations. Current methods often overlook fine-grained information across multi-magnification WSIs and variations in tumor microenvironments. Here, we propose an easy-to-hard progressive representation learning, termed dual-curriculum contrastive multi-instance learning (DCMIL), to efficiently process WSIs for cancer prognosis. The model does not rely on dense annotations and enables the direct transformation of gigapixel-size WSIs into outcome predictions. Extensive experiments on twelve cancer types (5,954 patients, 12.54 million tiles) demonstrate that DCMIL outperforms standard WSI-based prognostic models. Additionally, DCMIL identifies fine-grained prognosis-salient regions, provides robust instance uncertainty estimation, and captures morphological differences between normal and tumor tissues, with the potential to generate new biological insights. All codes have been made publicly accessible at https://github.com/tuuuc/DCMIL.

cs.CV↗

arXiv2025

Hybrid Advertising in the Sponsored Search

Online advertisements are a primary revenue source for e-commerce platforms. Traditional advertising models are store-centric, selecting winning stores through auction mechanisms. Recently, a new approach known as joint advertising has emerged, which presents sponsored bundles combining one store and one brand in ad slots. Unlike traditional models, joint advertising allows platforms to collect payments from both brands and stores. However, each of these two advertising models appeals to distinct user groups, leading to low click-through rates when users encounter an undesirable advertising model. To address this limitation and enhance generality, we propose a novel advertising model called ''Hybrid Advertising''. In this model, each ad slot can be allocated to either an independent store or a bundle. To find the optimal auction mechanisms in hybrid advertising, while ensuring nearly dominant strategy incentive compatibility and individual rationality, we introduce the Hybrid Regret Network (HRegNet), a neural network architecture designed for this purpose. Extensive experiments on both synthetic and real-world data demonstrate that the mechanisms generated by HRegNet significantly improve platform revenue compared to established baseline methods.

cs.GT↗

arXiv2025

Matching-Free Depth Recovery from Structured Light

We introduce a novel approach for depth estimation using images obtained from monocular structured light systems. In contrast to many existing methods that depend on image matching, our technique employs a density voxel grid to represent scene geometry. This grid is trained through self-supervised differentiable volume rendering. Our method leverages color fields derived from the projected patterns in structured light systems during the rendering process, facilitating the isolated optimization of the geometry field. This innovative approach leads to faster convergence and high-quality results. Additionally, we integrate normalized device coordinates (NDC), a distortion loss, and a distinctive surface-based color loss to enhance geometric fidelity. Experimental results demonstrate that our method outperforms current matching-based techniques in terms of geometric performance in few-shot scenarios, achieving an approximately 30% reduction in average estimated depth errors for both synthetic scenes and real-world captured scenes. Moreover, our approach allows for rapid training, being approximately three times faster than previous matching-free methods that utilize implicit representations.

cs.CV↗

arXiv2025

APTOS-2024 challenge report: Generation of synthetic 3D OCT images from fundus photographs

Optical Coherence Tomography (OCT) provides high-resolution, 3D, and non-invasive visualization of retinal layers in vivo, serving as a critical tool for lesion localization and disease diagnosis. However, its widespread adoption is limited by equipment costs and the need for specialized operators. In comparison, 2D color fundus photography offers faster acquisition and greater accessibility with less dependence on expensive devices. Although generative artificial intelligence has demonstrated promising results in medical image synthesis, translating 2D fundus images into 3D OCT images presents unique challenges due to inherent differences in data dimensionality and biological information between modalities. To advance generative models in the fundus-to-3D-OCT setting, the Asia Pacific Tele-Ophthalmology Society (APTOS-2024) organized a challenge titled Artificial Intelligence-based OCT Generation from Fundus Images. This paper details the challenge framework (referred to as APTOS-2024 Challenge), including: the benchmark dataset, evaluation methodology featuring two fidelity metrics-image-based distance (pixel-level OCT B-scan similarity) and video-based distance (semantic-level volumetric consistency), and analysis of top-performing solutions. The challenge attracted 342 participating teams, with 42 preliminary submissions and 9 finalists. Leading methodologies incorporated innovations in hybrid data preprocessing or augmentation (cross-modality collaborative paradigms), pre-training on external ophthalmic imaging datasets, integration of vision foundation models, and model architecture improvement. The APTOS-2024 Challenge is the first benchmark demonstrating the feasibility of fundus-to-3D-OCT synthesis as a potential solution for improving ophthalmic care accessibility in under-resourced healthcare settings, while helping to expedite medical research and clinical applications.

cs.CV↗