arXiv subjects

Chuanfu Xu

Publications and source records attributed to Chuanfu Xu.

11 recordsLinked to original sources

arXiv2026

CW-Ghost: Search-Free Granularity Selection for Helper-Thread Prefetching via Capacity Windows

Helper-thread prefetching hides the latency of irregular memory accesses by executing address dependency chains ahead of the main thread. However, its effectiveness depends on the range of future iterations covered by the helper thread. A fixed coverage range cannot consistently accommodate different workloads and processors, whereas exhaustively evaluating candidate configurations incurs substantial configuration cost. This paper presents CW-Ghost, which uses a single offline profiling run to estimate the average demand cache line fill volume generated per target iteration in a target region. CW-Ghost combines this estimate with a cache capacity budget to derive a Capacity Window, which determines the iteration granularity of each prefetch chunk. In addition, bounded chunk-level synchronization limits the number of chunks by which the helper thread may run ahead of the main thread. Across 14 workload instances evaluated on Intel and AMD CPU platforms, CW-Ghost achieves geometric mean speedups of 1.54x and 1.33x, respectively, over the original programs. Compared with Ghost Threading, it improves geometric mean performance by 15.8% and 10.8%, respectively, while achieving more than 99% of the empirically optimal performance within the candidate set on both platforms. These results demonstrate that cache capacity constraints can effectively guide the selection of granularity for helper-thread prefetching.

cs.DC↗

arXiv2026

FreeSpec: Training-Free Long Video Generation via Singular-Spectrum Reconstruction

Video diffusion models perform well in short-video synthesis, but their training-free extension to long videos often suffers from content drift, temporal inconsistency, and over-smoothed dynamics. Existing methods improve temporal consistency by combining a global branch with a local branch, but they often further decompose appearance consistency and temporal dynamics within each branch using predefined criteria. This assignment is unreliable when appearance and action progression are tightly coupled, such as in camera motion and sequential motion. We analyze the video temporal extension issue from a singular-spectrum perspective and show that enlarged self-attention windows induce spectral concentration: spectral energy becomes dominated by a few low-rank singular directions, preserving coarse structure but suppressing high-rank spatial details and motion-rich temporal variations. To mitigate this problem, we propose FreeSpec, a training-free spectral reconstruction framework for long-video generation. FreeSpec decomposes global and local features with singular value decomposition, and uses the global branch as low-rank spectral guidance and the local branch as a high-rank reconstruction basis. This spectrum-level fusion avoids the rigid feature partitioning of previous decomposition rules, preserving long-range consistency while better retaining spatial details and temporal dynamics. Experiments on Wan2.1 and LTX-Video demonstrate that FreeSpec improves long-video generation, especially for temporal dynamics, while maintaining strong visual quality and temporal consistency. Project demo: https://fdchen24.github.io/FreeSpec-Website/.

cs.CV↗

arXiv2025

JointTuner: Appearance-Motion Adaptive Joint Training for Customized Video Generation

Recent advancements in customized video generation have led to significant improvements in the simultaneous adaptation of appearance and motion. Typically, decoupling the appearance and motion training, prior methods often introduce concept interference, resulting in inaccurate rendering of appearance features or motion patterns. In addition, these methods often suffer from appearance contamination, in which background and foreground elements from reference videos distort the customized video. This paper aims to alleviate these issues by proposing JointTuner. The core motivation of our JointTuner is to enable joint optimization of both appearance and motion components, upon which two key innovations are developed, i.e., Gated Low-Rank Adaptation (GLoRA) and Appearance-independent Temporal Loss (AiT Loss). Specifically, GLoRA uses a context-aware activation layer, analogous to a gating regulator, to dynamically steer LoRA modules toward learning either appearance or motion while maintaining spatio-temporal consistency. Moreover, with the finding that channel-temporal shift noise suppresses appearance-related low-frequencies while enhancing motion-related high-frequencies, we designed the AiT Loss. This loss adds the same shift to the diffusion model's predicted noise during fine-tuning, forcing the model to prioritize learning motion patterns. JointTuner's architecture-agnostic design supports both UNet (e.g., ZeroScope) and Diffusion Transformer (e.g., CogVideoX) backbones, ensuring its customization capabilities scale with the evolution of foundational video models. Furthermore, we present a systematic evaluation framework for appearance-motion combined customization, covering 90 combinations evaluated along four critical dimensions: semantic alignment, motion dynamism, temporal consistency, and perceptual quality. Our project homepage is available online.

cs.CV↗

arXiv2025

"Stones from Other Hills can Polish Jade": Zero-shot Anomaly Image Synthesis via Cross-domain Anomaly Injection

Industrial image anomaly detection (IAD) is a pivotal topic with huge value. Due to anomaly's nature, real anomalies in a specific modern industrial domain (i.e. domain-specific anomalies) are usually too rare to collect, which severely hinders IAD. Thus, zero-shot anomaly synthesis (ZSAS), which synthesizes pseudo anomaly images without any domain-specific anomaly, emerges as a vital technique for IAD. However, existing solutions are either unable to synthesize authentic pseudo anomalies, or require cumbersome training. Thus, we focus on ZSAS and propose a brand-new paradigm that can realize both authentic and training-free ZSAS. It is based on a chronically-ignored fact: Although domain-specific anomalies are rare, real anomalies from other domains (i.e. cross-domain anomalies) are actually abundant and directly applicable to ZSAS. Specifically, our new ZSAS paradigm makes three-fold contributions: First, we propose a novel method named Cross-domain Anomaly Injection (CAI), which directly exploits cross-domain anomalies to enable highly authentic ZSAS in a training-free manner. Second, to supply CAI with sufficient cross-domain anomalies, we build the first Domain-agnostic Anomaly Dataset within our best knowledge, which provides ZSAS with abundant real anomaly patterns. Third, we propose a CAI-guided Diffusion Mechanism, which further breaks the quantity limit of real anomalies and enable unlimited anomaly synthesis. Our head-to-head comparison with existing ZSAS solutions justifies our paradigm's superior performance for IAD and demonstrates it as an effective and pragmatic ZSAS solution.

cs.CV↗

arXiv2025

Mitigating Sensitive Information Leakage in LLMs4Code through Machine Unlearning

Large Language Models for Code (LLMs4Code) have achieved strong performance in code generation, but recent studies reveal that they may memorize and leak sensitive information contained in training data, posing serious privacy risks. To address this gap, this work presents the first comprehensive empirical study on applying machine unlearning to mitigate sensitive information leakage in LLMs4Code. We first construct a dedicated benchmark that includes: (i) a synthetic forget set containing diverse forms of personal information, and (ii) a retain set designed to evaluate whether code-generation capability is preserved after unlearning. Using this benchmark, we systematically assess three representative unlearning algorithms (GA, GA+GD, GA+KL) across three widely used open-source LLMs4Code models (AIXCoder-7B, CodeLlama-7B, CodeQwen-7B). Experimental results demonstrate that machine unlearning can substantially reduce direct memorization-based leakage: on average, the direct leak rate drops by more than 50% while retaining about over 91% of the original code-generation performance. Moreover, by analyzing post-unlearning outputs, we uncover a consistent shift from direct to indirect leakage, revealing an underexplored vulnerability that persists even when the target data has been successfully forgotten. Our findings show that machine unlearning is a feasible and effective solution for enhancing privacy protection in LLMs4Code, while also highlighting the need for future techniques capable of mitigating both direct and indirect leakage simultaneously.

cs.CR↗

arXiv2022

An Unsupervised Short- and Long-Term Mask Representation for Multivariate Time Series Anomaly Detection

Anomaly detection of multivariate time series is meaningful for system behavior monitoring. This paper proposes an anomaly detection method based on unsupervised Short- and Long-term Mask Representation learning (SLMR). The main idea is to extract short-term local dependency patterns and long-term global trend patterns of the multivariate time series by using multi-scale residual dilated convolution and Gated Recurrent Unit(GRU) respectively. Furthermore, our approach can comprehend temporal contexts and feature correlations by combining spatial-temporal masked self-supervised representation learning and sequence split. It considers the importance of features is different, and we introduce the attention mechanism to adjust the contribution of each feature. Finally, a forecasting-based model and a reconstruction-based model are integrated to focus on single timestamp prediction and latent representation of time series. Experiments show that the performance of our method outperforms other state-of-the-art models on three real-world datasets. Further analysis shows that our method is good at interpretability.

cs.LG↗

arXiv2022

Deep Anomaly Discovery From Unlabeled Videos via Normality Advantage and Self-Paced Refinement

While classic video anomaly detection (VAD) requires labeled normal videos for training, emerging unsupervised VAD (UVAD) aims to discover anomalies directly from fully unlabeled videos. However, existing UVAD methods still rely on shallow models to perform detection or initialization, and they are evidently inferior to classic VAD methods. This paper proposes a full deep neural network (DNN) based solution that can realize highly effective UVAD. First, we, for the first time, point out that deep reconstruction can be surprisingly effective for UVAD, which inspires us to unveil a property named "normality advantage", i.e., normal events will enjoy lower reconstruction loss when DNN learns to reconstruct unlabeled videos. With this property, we propose Localization based Reconstruction (LBR) as a strong UVAD baseline and a solid foundation of our solution. Second, we propose a novel self-paced refinement (SPR) scheme, which is synthesized into LBR to conduct UVAD. Unlike ordinary self-paced learning that injects more samples in an easy-to-hard manner, the proposed SPR scheme gradually drops samples so that suspicious anomalies can be removed from the learning process. In this way, SPR consolidates normality advantage and enables better UVAD in a more proactive way. Finally, we further design a variant solution that explicitly takes the motion cues into account. The solution evidently enhances the UVAD performance, and it sometimes even surpasses the best classic VAD methods. Experiments show that our solution not only significantly outperforms existing UVAD methods by a wide margin (5% to 9% AUROC), but also enables UVAD to catch up with the mainstream performance of classic VAD.

cs.CV↗

arXiv2020

Cloze Test Helps: Effective Video Anomaly Detection via Learning to Complete Video Events

As a vital topic in media content interpretation, video anomaly detection (VAD) has made fruitful progress via deep neural network (DNN). However, existing methods usually follow a reconstruction or frame prediction routine. They suffer from two gaps: (1) They cannot localize video activities in a both precise and comprehensive manner. (2) They lack sufficient abilities to utilize high-level semantics and temporal context information. Inspired by frequently-used cloze test in language study, we propose a brand-new VAD solution named Video Event Completion (VEC) to bridge gaps above: First, we propose a novel pipeline to achieve both precise and comprehensive enclosure of video activities. Appearance and motion are exploited as mutually complimentary cues to localize regions of interest (RoIs). A normalized spatio-temporal cube (STC) is built from each RoI as a video event, which lays the foundation of VEC and serves as a basic processing unit. Second, we encourage DNN to capture high-level semantics by solving a visual cloze test. To build such a visual cloze test, a certain patch of STC is erased to yield an incomplete event (IE). The DNN learns to restore the original video event from the IE by inferring the missing patch. Third, to incorporate richer motion dynamics, another DNN is trained to infer erased patches' optical flow. Finally, two ensemble strategies using different types of IE and modalities are proposed to boost VAD performance, so as to fully exploit the temporal context and modality information for VAD. VEC can consistently outperform state-of-the-art methods by a notable margin (typically 1.5%-5% AUROC) on commonly-used VAD benchmarks. Our codes and results can be verified at github.com/yuguangnudt/VEC_VAD.

cs.CV↗

arXiv2019

Characterizing Scalability of Sparse Matrix-Vector Multiplications on Phytium FT-2000+ Many-cores

Understanding the scalability of parallel programs is crucial for software optimization and hardware architecture design. As HPC hardware is moving towards many-core design, it becomes increasingly difficult for a parallel program to make effective use of all available processor cores. This makes scalability analysis increasingly important. This paper presents a quantitative study for characterizing the scalability of sparse matrix-vector multiplications (SpMV) on Phytium FT-2000+, an ARM-based many-core architecture for HPC computing. We choose to study SpMV as it is a common operation in scientific and HPC applications. Due to the newness of ARM-based many-core architectures, there is little work on understanding the SpMV scalability on such hardware design. To close the gap, we carry out a large-scale empirical evaluation involved over 1,000 representative SpMV datasets. We show that, while many computation-intensive SpMV applications contain extensive parallelism, achieving a linear speedup is non-trivial on Phytium FT-2000+. To better understand what software and hardware parameters are most important for determining the scalability of a given SpMV kernel, we develop a performance analytical model based on the regression tree. We show that our model is highly effective in characterizing SpMV scalability, offering useful insights to help application developers for better optimizing SpMV on an emerging HPC architecture.

cs.DC↗

arXiv2018

Optimizing Sparse Matrix-Vector Multiplication on Emerging Many-Core Architectures

Sparse matrix vector multiplication (SpMV) is one of the most common operations in scientific and high-performance applications, and is often responsible for the application performance bottleneck. While the sparse matrix representation has a significant impact on the resulting application performance, choosing the right representation typically relies on expert knowledge and trial and error. This paper provides the first comprehensive study on the impact of sparse matrix representations on two emerging many-core architectures: the Intel's Knights Landing (KNL) XeonPhi and the ARM-based FT-2000Plus (FTP). Our large-scale experiments involved over 9,500 distinct profiling runs performed on 956 sparse datasets and five mainstream SpMV representations. We show that the best sparse matrix representation depends on the underlying architecture and the program input. To help developers to choose the optimal matrix representation, we employ machine learning to develop a predictive model. Our model is first trained offline using a set of training examples. The learned model can be used to predict the best matrix representation for any unseen input for a given architecture. We show that our model delivers on average 95% and 91% of the best available performance on KNL and FTP respectively, and it achieves this with no runtime profiling overhead.

cs.MS↗

arXiv2013

An Empirical Study of Intel Xeon Phi

With at least 50 cores, Intel Xeon Phi is a true many-core architecture. Featuring fairly powerful cores, two cache levels, and very fast interconnections, the Xeon Phi can get a theoretical peak of 1000 GFLOPs and over 240 GB/s. These numbers, as well as its flexibility - it can be used both as a coprocessor or as a stand-alone processor - are very tempting for parallel applications looking for new performance records. In this paper, we present an empirical study of Xeon Phi, stressing its performance limits and relevant performance factors, ultimately aiming to present a simplified view of the machine for regular programmers in search for performance. To do so, we have micro-benchmarked the main hardware components of the processor - the cores, the memory hierarchies, the ring interconnect, and the PCIe connection. We show that, in ideal microbenchmarking conditions, the performance that can be achieved is very close to the theoretical peak, as given in the official programmer's guide. We have also identified and quantified several causes for significant performance penalties. Our findings have been captured in four optimization guidelines, and used to build a simplified programmer's view of Xeon Phi, eventually enable the design and prototyping of applications on a functionality-based model of the architecture.

cs.DC↗