Searcharxiv⌕ Search

arXiv subjects

Wenbin Luo

Publications and source records attributed to Wenbin Luo.

12 recordsLinked to original sources
arXiv2026

TAGR: Temporally Adaptive Generative Recommendation for Industrial Live-Streaming Advertising

Live-streaming advertising is an important monetization channel on short-video and e-commerce platforms, where rapidly changing live content, promoted products, and user feedback impose strong freshness requirements on recommendation models. Existing generative recommenders designed for static domains fail at three levels: static semantic IDs (SID) cannot track evolving live ads; single-scale behavior modeling misses shifting intent; preference optimization conflicts between fresh on-policy feedback and training stability. We propose TAGR, a generative recommendation framework with temporal adaptation at three levels: live-ad tokenization, user intent modeling, and preference alignment. At the token level, Live Semantic-Collaborative ID (LSID) periodically refreshes each active ad's SID based on its current live scene and promoted products, while retaining a stable hierarchical token vocabulary for autoregressive generation. At the intent level, Intent-Aware Generation (IAG) models live-room entry histories at multiple temporal granularities as the primary intent sequence, keeps auxiliary behaviors as separate inputs, and weights next-token prediction (NTP) using post-request intent evidence and business value. At the alignment level, Intermittent On-Policy Preference Optimization (IOPO) periodically samples fresh candidate groups from the current policy and performs behavior- and value-aligned preference updates interleaved with supervised NTP maintenance to preserve learned behavior distribution. Deployed on a large-scale e-commerce live-stream advertising platform, TAGR improves live-room entry and shopping-cart click rates by 8.5% and 7.4%, respectively, and achieves a 16.1% revenue lift over the production baseline. These results demonstrate the effectiveness and industrial viability of temporally adaptive generative recommendation for live-stream advertising.

cs.IR↗
arXiv2025

Online Continual Learning via Spiking Neural Networks with Sleep Enhanced Latent Replay

Edge computing scenarios necessitate the development of hardware-efficient online continual learning algorithms to be adaptive to dynamic environment. However, existing algorithms always suffer from high memory overhead and bias towards recently trained tasks. To tackle these issues, this paper proposes a novel online continual learning approach termed as SESLR, which incorporates a sleep enhanced latent replay scheme with spiking neural networks (SNNs). SESLR leverages SNNs' binary spike characteristics to store replay features in single bits, significantly reducing memory overhead. Furthermore, inspired by biological sleep-wake cycles, SESLR introduces a noise-enhanced sleep phase where the model exclusively trains on replay samples with controlled noise injection, effectively mitigating classification bias towards new classes. Extensive experiments on both conventional (MNIST, CIFAR10) and neuromorphic (NMNIST, CIFAR10-DVS) datasets demonstrate SESLR's effectiveness. On Split CIFAR10, SESLR achieves nearly 30% improvement in average accuracy with only one-third of the memory consumption compared to baseline methods. On Split CIFAR10-DVS, it improves accuracy by approximately 10% while reducing memory overhead by a factor of 32. These results validate SESLR as a promising solution for online continual learning in resource-constrained edge computing scenarios.

cs.NE↗
arXiv2025

Geometric Bogomolov conjecture for semiabelian varieties

We establish the geometric Bogomolov conjecture for semiabelian varieties over function fields. We show a closed subvariety contains Zariski dense sets of small points, if and only if, after modulo its stabilizer, it is a torsion translate of a constant variety. A new phenomenon is that a special subvariety may not have a Zariski dense set of points of height 0.

math.AG↗
arXiv2024

Arakelov geometry on flag varieties over function fields and related topics

Let $k$ be an algebraically closed field of characteristic zero. Let $G$ be a connected reductive group over $k$, $P \subseteq G$ be a parabolic subgroup and $\lambda: P \longrightarrow G$ be a strictly anti-dominant character. Let $C$ be a projective smooth curve over $k$ with function field $K=k(C)$ and $F$ be a principal $G$-bundle on $C$. Then $F/P \longrightarrow C$ is a flag bundle and $\mathcal{L}_\lambda=F \times_P k_\lambda$ on $F/P$ is a relatively ample line bundle. We compute the height filtration, successive minima, and the Boucksom-Chen concave transform of the height function $h_{\mathcal{L}_\lambda}: X(\overline{K}) \longrightarrow \mathbb{R}$ over the flag variety $X=(F/P)_K$. An interesting application is that the height of $X$ equals to a weighted average of successive minima, and one may view this as a refinement of Zhang's inequality of successive minima. Let $f \in N^1(F/P)$ be the numerical class of a vertical fiber. We compute the augmented base loci $\mathrm{B}_+(\mathcal{L}_\lambda-tf)$ for any $t \in \mathbb{R}$, and it turns out that they are almost the same as the height filtration. As a corollary, we compute the $k$-th movable cones of flag bundles over curves for all $k$.

math.NT↗
arXiv2024

Arithmetic degrees of dynamical systems over fields of characteristic zero

In this article, we generalize the arithmetic degree and its related theory to dynamical systems defined over an arbitrary field $\mathbf{k}$ of characteristic $0$. We first consider a dynamical system $(X,f)$ over a finitely generated field $K$ over $\mathbb{Q}$, we introduce the arithmetic degrees $\alpha(f,\cdot)$ for $\overline{K}$-points by using Moriwaki heights. We study the arithmetic dynamical degree of $(X,f)$ and establish the relative degree formula. The relative degree formula gives a proof of the fundamental inequality, that is, the upper arithmetic degree $\overline{\alpha}(f,x)$ is less than or equal to the first dynamical degree $\lambda_1(f)$ in this setting. By taking spread-outs, we extend the definition of arithmetic degrees to dynamical systems over the field $\mathbf k$. We demonstrate that our definition is independent of the choice of the spread-out. Moreover, in this setting, we prove certain special cases of the Kawaguchi-Silverman conjecture. A main novelty of this paper is that, we give a characterization of arithmetic degrees of "transcendental points" in the case $\mathbf{k}=\mathbb{C}$, from which we deduce that $\alpha(f,x)=\lambda_1(f)$ for very general $x\in X(\mathbb{C})$ when $f$ is an endomorphism.

math.NT↗
arXiv2023

Vector-based Efficient Data Hiding in Encrypted Images via Multi-MSB Replacement

As an essential technique for data privacy protection, reversible data hiding in encrypted images (RDHEI) methods have drawn intensive research interest in recent years. In response to the increasing demand for protecting data privacy, novel methods that perform RDHEI are continually being developed. We propose two effective multi-MSB (most significant bit) replacement-based approaches that yield comparably high data embedding capacity, improve overall processing speed, and enhance reconstructed images' quality. Our first method, Efficient Multi-MSB Replacement-RDHEI (EMR-RDHEI), obtains higher data embedding rates (DERs, also known as payloads) and better visual quality in reconstructed images when compared with many other state-of-the-art methods. Our second method, Lossless Multi-MSB Replacement-RDHEI (LMR-RDHEI), can losslessly recover original images after an information embedding process is performed. To verify the accuracy of our methods, we compared them with other recent RDHEI techniques and performed extensive experiments using the widely accepted BOWS-2 dataset. Our experimental results showed that the DER of our EMR-RDHEI method ranged from 1.2087 bit per pixel (bpp) to 6.2682 bpp with an average of 3.2457 bpp. For the LMR-RDHEI method, the average DER was 2.5325 bpp, with a range between 0.2129 bpp and 6.0168 bpp. Our results demonstrate that these methods outperform many other state-of-the-art RDHEI algorithms. Additionally, the multi-MSB replacement-based approach provides a clean design and efficient vectorized implementation.

cs.CR↗
arXiv2022

Slope boundedness and Equidistribution theorem

In this article, we prove the boundedness of minimal slopes of adelic line bundles over function fields of characteristic 0. This can be applied to prove the equidistribution of generic and small points with respect to a big and semipositive adelic line bundle. Our methods can be applied to the finite places of number fields as well. We also show the continuity of $\chi$-volumes over function fields.

math.NT↗
arXiv2022

A relative bigness inequality and equidistribution theorem over function fields

For any line bundle written as a subtraction of two ample line bundles, Siu's inequality gives a criterion on its bigness. We generalize this inequality to a relative case. The arithmetic meaning behind the inequality leads to its application on algebraic dynamic systems, which is the equidistribution theorem of generic and small net of subvarieties over a function field.

math.AG↗
arXiv2020

The continuity of $χ$-volume functions over adelic curves

In the setting of Arakelov geometry over adelic curves, we introduce the $χ$-volume function and show some general properties. This article is dedicated to talk about the continuity of $χ$-volume function. By discussing its relationship with volume function, we prove its continuity around adelic $\mathbb{Q}$-ample $\mathbb{Q}$-Cartier divisors and its continuity in the trivially valued case. The study of the variation of arithmetic Okounkov bodies leads us to its continuous extension on arithmetic surfaces.

math.AG↗
arXiv2020

Arithmetic over trivially valued field and its applications

By some result on the study of arithemtic over trivially valued field, we find its applications to Arakelov geometry over adelic curves. We prove a partial result of the continuity of arithmetic $χ$-volume along semiample divisors. Moreover, we give a upper bound estimate of arithmetic Hilbert-Samuel function.

math.AG↗
arXiv2020

Application of Structural Similarity Analysis of Visually Salient Areas and Hierarchical Clustering in the Screening of Similar Wireless Capsule Endoscopic Images

Small intestinal capsule endoscopy is the mainstream method for inspecting small intestinal lesions,but a single small intestinal capsule endoscopy will produce 60,000 - 120,000 images, the majority of which are similar and have no diagnostic value. It takes 2 - 3 hours for doctors to identify lesions from these images. This is time-consuming and increase the probability of misdiagnosis and missed diagnosis since doctors are likely to experience visual fatigue while focusing on a large number of similar images for an extended period of time.In order to solve these problems, we proposed a similar wireless capsule endoscope (WCE) image screening method based on structural similarity analysis and the hierarchical clustering of visually salient sub-image blocks. The similarity clustering of images was automatically identified by hierarchical clustering based on the hue,saturation,value (HSV) spatial color characteristics of the images,and the keyframe images were extracted based on the structural similarity of the visually salient sub-image blocks, in order to accurately identify and screen out similar small intestinal capsule endoscopic images. Subsequently, the proposed method was applied to the capsule endoscope imaging workstation. After screening out similar images in the complete data gathered by the Type I OMOM Small Intestinal Capsule Endoscope from 52 cases covering 17 common types of small intestinal lesions, we obtained a lesion recall of 100% and an average similar image reduction ratio of 76%. With similar images screened out, the average play time of the OMOM image workstation was 18 minutes, which greatly reduced the time spent by doctors viewing the images.

eess.IV↗