arXiv subjects

Bowen Xue

Publications and source records attributed to Bowen Xue.

16 recordsLinked to original sources

arXiv2026

Ring Forcing: Towards Precise Long-Term Memory for Autoregressive Video Diffusion

Scaling video generation to long durations reveals a critical bottleneck: current models lack robust long-term memory. This deficiency can be studied along two critical aspects: object permanence, the ability to precisely reproduce the appearance of objects upon re-entry; and memory capacity, the ability to process ultra-long context and use information from distant history. Robust long-term memory requires both: object permanence without sufficient context handling limits the temporal scope, while long context length without permanence fails to maintain identity. To address this, we present Ring Forcing, an autoregressive video diffusion framework designed to robustly construct and precisely utilize long-term memory. Our ring-structured training strategy enforces retrieval from distant history, effectively reconciling the trade-off between strict historical adherence and generative diversity. To expand memory capacity, we introduce a compression and timestep composition strategy. Under fixed sequence length constraints, this method extends the effective historical span to minutes-long durations and achieves a comprehensive receptive field over the entire history. Furthermore, we present a sparse RoPE mechanism to enable flexible, scalable memory adaptation while fully exploiting pre-trained priors. Extensive experiments demonstrate that Ring Forcing achieves superior minutes-long coherence and object permanence, significantly outperforming state-of-the-art methods.

cs.CV↗

arXiv2026

Latent-Frequency Validity: Fast Spectral Editing with Screened Video-VAE Transfer Operators

Direct spectral editing in video-VAE latents can control noise, flicker, smoothness, and frequency content without a decode--filter--reencode pass. However, video VAEs may redistribute pixel-space frequency bands across latent channels, and latent edits can disrupt VAE round-trip dynamics. We introduce \emph{latent-frequency validity} (LFV), which learns a compact VAE-specific spectral response and deploys it only when it improves decoded-target fidelity without worsening round-trip drift. LFV follows a validation-selected path from a diagonal per-frequency calibrator (C1) to full channel mixing (CM), making cross-channel capacity a controllable per-edit resource. Across 544 VAE--edit cells spanning six spectral families, LFV emits 423 cheap operators: 277 are handled by C1, while 146 (34.5\% of emitted operators) require channel mixing. On the primary 120-cell radial sweep, 99/100 emitted operators pass source-video-grouped held-out evaluation. Across five additional filter families, all 323 emitted operators pass held-out evaluation. Fully frozen OpenVid-fitted operators, including the validation-selected path coefficient, pass all 20 tested CogVideoX and HunyuanVideo generated-domain cells without adaptation. The selected response matches direct latent-filter latency and is about $3\times$ faster than pixel filter--reencode. The resulting maps reveal distinct VAE regimes, including strongly channel-coupled CogVideoX responses and a sharp Open-Sora high-band stability frontier.

cs.CV↗

arXiv2026

FourTune: Towards Fully 4-Bit Efficient Post-Training for Diffusion Models

Diffusion models have become a dominant paradigm for high-quality generative modeling, while post-training is essential for adapting them to diverse downstream applications. However, post-training of large diffusion models is still challenging due to the prohibitive memory footprints and slow training speed, which existing parameter-efficient fine-tuning methods only partially address. To overcome these limitations, we propose FourTune, an efficient post-training framework for diffusion models based on an end-to-end W4A4G4 paradigm. FourTune introduces a triple-branch hybrid pipeline that augments the standard LoRA architecture with a frozen numerical stabilizer to isolate quantization-sensitive outliers, enabling stable training under native 4-bit computation. In addition, FourTune employs hardware-efficient block-wise quantization and customized fused kernels to support efficient quantized backpropagation and reduce memory bandwidth overhead. Across customization, reinforcement learning, and distillation tasks, FourTune matches the quality of full-precision fine-tuning. On FLUX.1-dev (12B), FourTune reduces memory overhead by 2.25$\times$ and increases end-to-end training throughput by 2.27$\times$ compared to BF16 LoRA.

cs.LG↗

arXiv2026

Secure AltDA Integration for Ethereum L2s: An End-to-End Validation Framework

Alternative data availability (AltDA) systems provide Ethereum L2s with an external data publication layer for high throughput rollup designs. By moving bulk data publication outside of Ethereum, AltDA allows L2s to process more data than native DA. However, this replacement introduces a new consensus critical integration layer. Existing ecosystem frameworks identify high level risks, such as external DA trust assumptions and the presence or absence of a DA verifier, but do not provide a complete specification for how an L2 should integrate with AltDA. This gap can lead to L2 halts, inconsistent derivation across honest L2 nodes, invalid state assertions, or bridge attacks. This paper presents a canonical validation framework for secure AltDA integration. We model the boundary as a typed, deterministic, and total translation from L1 inbox bytes to an AltDA commitment, then to externally available data, and finally to the rollup payload consumed by the rest of core L2s logic. The central principle is that every adversarial input must lead to a defined unique outcome. We show how missing obligations lead to concrete failure modes, including underconstrained settlement, derivation halts, inconsistent honest node behavior, invalid state assertions, and bridge safety failures. We then apply the framework to representative AltDA integration architectures, including Celestia-Blobstream, EigenDA based designs, and Avail-ZKsync. Our evaluation shows that secure AltDA integration is not determined solely by the DA provider or bridge. The surrounding L2 integration must also enforce the full validation relation connecting L1 inbox inputs to accepted L2 state.

cs.CR↗

arXiv2026

PBR-Inspired Controllable Diffusion for Image Generation

Despite recent advances in text-to-image generation, controlling geometric layout and PBR material properties in synthesized scenes remains challenging. We present a pipeline that first produces a G-buffer (albedo, normals, depth, roughness, shading, and metallic) from a text prompt and then renders a final image through a PBR-inspired branch network. This intermediate representation enables fine-grained control: users can copy and paste within specific G-buffer channels to insert or reposition objects, or apply masks to the irradiance channel to adjust lighting locally. As a result, real objects can be seamlessly integrated into virtual scenes. By separating user-friendly scene description from image rendering, our method offers a practical balance between detailed post-generation control and efficient text-driven synthesis. We demonstrate its effectiveness through quantitative evaluations and a user study with 156 participants, showing consistent human preference over strong baselines and confirming that G-buffer control extends the flexibility of text-guided image generation.

cs.GR↗

arXiv2026

VideoNeuMat: Neural Material Extraction from Generative Video Models

Creating photorealistic materials for 3D rendering requires exceptional artistic skill. Generative models for materials could help, but are currently limited by the lack of high-quality training data. While recent video generative models effortlessly produce realistic material appearances, this knowledge remains entangled with geometry and lighting. We present VideoNeuMat, a two-stage pipeline that extracts reusable neural material assets from video diffusion models. First, we finetune a large video model (Wan 2.1 14B) to generate material sample videos under controlled camera and lighting trajectories, effectively creating a "virtual gonioreflectometer" that preserves the model's material realism while learning a structured measurement pattern. Second, we reconstruct compact neural materials from these videos through a Large Reconstruction Model (LRM) finetuned from a smaller Wan 1.3B video backbone. From 17 generated video frames, our LRM performs single-pass inference to predict neural material parameters that generalize to novel viewing and lighting conditions. The resulting materials exhibit realism and diversity far exceeding the limited synthetic training data, demonstrating that material knowledge can be successfully transferred from internet-scale video models into standalone, reusable neural 3D assets.

cs.CV↗

arXiv2025

OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search

Traditional e-commerce search systems employ multi-stage cascading architectures (MCA) that progressively filter items through recall, pre-ranking, and ranking stages. While effective at balancing computational efficiency with business conversion, these systems suffer from fragmented computation and optimization objective collisions across stages, which ultimately limit their performance ceiling. To address these, we propose \textbf{OneSearch}, the first industrial-deployed end-to-end generative framework for e-commerce search. This framework introduces three key innovations: (1) a Keyword-enhanced Hierarchical Quantization Encoding (KHQE) module, to preserve both hierarchical semantics and distinctive item attributes while maintaining strong query-item relevance constraints; (2) a multi-view user behavior sequence injection strategy that constructs behavior-driven user IDs and incorporates both explicit short-term and implicit long-term sequences to model user preferences comprehensively; and (3) a Preference-Aware Reward System (PARS) featuring multi-stage supervised fine-tuning and adaptive reward-weighted ranking to capture fine-grained user preferences. Extensive offline evaluations on large-scale industry datasets demonstrate OneSearch's superior performance for high-quality recall and ranking. The rigorous online A/B tests confirm its ability to enhance relevance in the same exposure position, achieving statistically significant improvements: +1.67% item CTR, +2.40% buyer, and +3.22% order volume. Furthermore, OneSearch reduces operational expenditure by 75.40% and improves Model FLOPs Utilization from 3.26% to 27.32%. The system has been successfully deployed across multiple search scenarios in Kuaishou, serving millions of users, generating tens of millions of PVs daily.

cs.IR↗

arXiv2025

Physics-Guided Motion Loss for Video Generation Model

Current video diffusion models generate visually compelling content but often violate basic laws of physics, producing subtle artifacts like rubber-sheet deformations and inconsistent object motion. We introduce a frequency-domain physics prior that improves motion plausibility without modifying model architectures. Our method decomposes common rigid motions (translation, rotation, scaling) into lightweight spectral losses, requiring only 2.7% of frequency coefficients while preserving 97%+ of spectral energy. Applied to Open-Sora, MVDIT, and Hunyuan, our approach improves both motion accuracy and action recognition by ~11% on average on OpenVID-1M (relative), while maintaining visual quality. User studies show 74--83% preference for our physics-enhanced videos. It also reduces warping error by 22--37% (depending on the backbone) and improves temporal consistency scores. These results indicate that simple, global spectral cues are an effective drop-in regularizer for physically plausible motion in video diffusion.

cs.CV↗

arXiv2025

Stand-In: A Lightweight and Plug-and-Play Identity Control for Video Generation

Generating high-fidelity human videos that match user-specified identities is important yet challenging in the field of generative AI. Existing methods often rely on an excessive number of training parameters and lack compatibility with other AIGC tools. In this paper, we propose Stand-In, a lightweight and plug-and-play framework for identity preservation in video generation. Specifically, we introduce a conditional image branch into the pre-trained video generation model. Identity control is achieved through restricted self-attentions with conditional position mapping. Thanks to these designs, which greatly preserve the pre-trained prior of the video generation model, our approach is able to outperform other full-parameter training methods in video quality and identity preservation, even with just $\sim$1% additional parameters and only 2000 training pairs. Moreover, our framework can be seamlessly integrated for other tasks, such as subject-driven video generation, pose-referenced video generation, stylization, and face swapping.

cs.CV↗

arXiv2024

ReflectanceFusion: Diffusion-based text to SVBRDF Generation

We introduce Reflectance Diffusion, a new neural text-to-texture model capable of generating high-fidelity SVBRDF maps from textual descriptions. Our method leverages a tandem neural approach, consisting of two modules, to accurately model the distribution of spatially varying reflectance as described by text prompts. Initially, we employ a pre-trained stable diffusion 2 model to generate a latent representation that informs the overall shape of the material and serves as our backbone model. Then, our ReflectanceUNet enables fine-tuning control over the material's physical appearance and generates SVBRDF maps. ReflectanceUNet module is trained on an extensive dataset comprising approximately 200,000 synthetic spatially varying materials. Our generative SVBRDF diffusion model allows for the synthesis of multiple SVBRDF estimates from a single textual input, offering users the possibility to choose the output that best aligns with their requirements. We illustrate our method's versatility by generating SVBRDF maps from a range of textual descriptions, both specific and broad. Our ReflectanceUNet model can integrate optional physical parameters, such as roughness and specularity, enhancing customization. When the backbone module is fixed, the ReflectanceUNet module refines the material, allowing direct edits to its physical attributes. Comparative evaluations demonstrate that ReflectanceFusion achieves better accuracy than existing text-to-material models, such as Text2Mat, while also providing the benefits of editable and relightable SVBRDF maps.

cs.GR↗

arXiv2024

A Hierarchical Architecture for Neural Materials

Neural reflectance models are capable of reproducing the spatially-varying appearance of many real-world materials at different scales. Unfortunately, existing techniques such as NeuMIP have difficulties handling materials with strong shadowing effects or detailed specular highlights. In this paper, we introduce a neural appearance model that offers a new level of accuracy. Central to our model is an inception-based core network structure that captures material appearances at multiple scales using parallel-operating kernels and ensures multi-stage features through specialized convolution layers. Furthermore, we encode the inputs into frequency space, introduce a gradient-based loss, and employ it adaptive to the progress of the learning phase. We demonstrate the effectiveness of our method using a variety of synthetic and real examples.

cs.GR↗

arXiv2024

Travelers: A scalable fair ordering BFT system

Many blockchain platform are subject to maximal value extraction (MEV), and users on the platform are losing money while sending transactions because the transaction order can be manipulated to extract value from them. Consensus protocols have been augmented with different notion of fair ordering in order to counter the problem. Out of all practical protocols, the most efficient BFT consensus requires $O(nTL + n^2T)$ communication complexity, where $n$ is number node, $T$ is number of transactions and $L$ is average transaction size. In this work, we propose a new system of BFT fair ordering protocols, Travelers, that substantially reduce the communication complexity. The proposed system of protocols satisfy a new notion of fair ordering, called probabilistic fair ordering, which is an extension to some existing notions of fairness. The new notion allows a small probability of error $ε$, that adversary can insert some transactions at any location in a block, but for the remaining $1-ε$ the a modified version of ordering linearizability holds. Our mechanism neither require a dissemination network nor direct submissions to all consensus nodes. The key innovation comes from a routing protocol, that is both flexible and efficient. We construct a protocol with $O(c\log({n})TL + n^2)$ communication complexity with $ε= 1/n^c$ for some system parameter $c\ge 1$.

cs.CR↗

arXiv2023

BigDipper: Sharded Censorship Resistant Data Availability for Leader-Based BFT

Leader-based Byzantine-fault-tolerant (BFT) protocols provide low latency and simple communication structure, but they give the leader short-term control over transaction inclusion. A malicious leader can keep the protocol live while delaying or excluding time-sensitive transactions such as auction bids, oracle updates, liquidations, and bridge messages. Existing responses often build a fixed censorship-resistance, hiding, or ordering mechanism into the protocol path, forcing all transactions to pay for the same protection level. name follows the end-to-end principle: the consensus layer exposes inclusion primitives rather than hardcoding stronger policies. Higher-layer protocols can then choose their own submission strategies and resources, whether through replication, erasure coding, or other mechanisms, to obtain the censorship-resistance, hiding, ordering, or execution guarantees they need. At the core of BigDipper is censorship-resistant data availability, or DA-CR, which certifies available replica-contributed mini-blocks for use by leader-based consensus. A central design goal is that data remains sharded on the consensus critical path: validators do not reconstruct or execute the full payload before voting, but instead check commitments, availability evidence, and the DA-CR inclusion rule. We define DA-CR guarantees for data-tampering resistance, honest mini-block inclusion, and residual leader influence. We then give concrete constructions based on erasure coding and linear commitments, analyze client-tunable transaction submission, and instantiate BigDipper inside HotStuff-2.

cs.CR↗

arXiv2023

Goldfish: Peer selection using Matrix completion in unstructured P2P network

Peer-to-peer (P2P) networks underlie a variety of decentralized paradigms including blockchains, distributed file storage and decentralized domain name systems. A central primitive in P2P networks is the peer selection algorithm, which decides how a node should select a fixed number of neighbors to connect with. In this paper, we consider the design of a peer-selection algorithm for unstructured P2P networks with the goal of minimizing the broadcast latency. We propose Goldfish, a novel solution that dynamically decides the neighbor set by exploiting the past experiences as well as exploring new neighbors. The key technical contributions come from bringing ideas of matrix completion for estimating message delivery times for every possible message for every peer ever connected, and a streaming algorithm to efficiently perform the estimation while achieving good performance. The matrix completion interpolates the delivery times to all virtual connections in order to select the best combination of neighbors. Goldfish employs a streaming algorithm that only uses a short recent memory to finish matrix interpolation. When the number of publishing source is equal to a node's maximal number of connections, Goldfish found the global optimal solution with 92.7% probability by exploring every node only once. In more complex situations where nodes are publishing based on exponential distribution and adjusting connection in real time, we compare Goldfish with a baseline peer selection system, and show Goldfish saves approximately 14.5% less time under real world geolocation and propagation latency.

cs.NI↗

arXiv2022

Dissecting Service Mesh Overheads

Service meshes play a central role in the modern application ecosystem by providing an easy and flexible way to connect different services that form a distributed application. However, because of the way they interpose on application traffic, they can substantially increase application latency and resource consumption. We develop a decompositional approach and a tool, called MeshInsight, to systematically characterize the overhead of service meshes and to help developers quantify overhead in deployment scenarios of interest. Using MeshInsight, we confirm that service meshes can have high overhead -- up to 185% higher latency and up to 92% more virtual CPU cores for our benchmark applications -- but the severity is intimately tied to how they are configured and the application workload. The primary contributors to overhead vary based on the configuration too. IPC (inter-process communication) and socket writes dominate when the service mesh operates as a TCP proxy, but protocol parsing dominates when it operates as an HTTP proxy. MeshInsight also enables us to study the end-to-end impact of optimizations to service meshes. We show that not all seemingly-promising optimizations lead to a notable overhead reduction in realistic settings.

cs.DC↗

arXiv2021

ACeD: Scalable Data Availability Oracle

A popular method in practice offloads computation and storage in blockchains by relying on committing only hashes of off-chain data into the blockchain. This mechanism is acknowledged to be vulnerable to a stalling attack: the blocks corresponding to the committed hashes may be unavailable at any honest node. The straightforward solution of broadcasting all blocks to the entire network sidesteps this data availability attack, but it is not scalable. In this paper, we propose ACeD, a scalable solution to this data availability problem with $O(1)$ communication efficiency, the first to the best of our knowledge. The key innovation is a new protocol that requires each of the $N$ nodes to receive only $O(1/N)$ of the block, such that the data is guaranteed to be available in a distributed manner in the network. Our solution creatively integrates coding-theoretic designs inside of Merkle tree commitments to guarantee efficient and tamper-proof reconstruction; this solution is distinct from Asynchronous Verifiable Information Dispersal (in guaranteeing efficient proofs of malformed coding) and Coded Merkle Tree (which only provides guarantees for random corruption as opposed to our guarantees for worst-case corruption). We implement ACeD with full functionality in 6000 lines of Rust code, integrate the functionality as a smart contract into Ethereum via a high-performance implementation demonstrating up to 10,000 transactions per second in throughput and 6000x reduction in gas cost on the Ethereum testnet Kovan.

cs.CR↗