检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Zhang, Mingyang Zhou, Yi Ren, Yi Zhang, Chen Yin, Xiang Li, Haizhou The Shenzhen Research Institute of Big Data School of Data Science The Chinese University of Hong Kong Shenzhen China The Department of Electrical and Computer Engineering National University of Singapore Singapore Speech & Audio Team ByteDance AI Lab

This paper proposes RefXVC, a method for cross-lingual voice conversion (XVC) that leverages reference information to improve conversion performance. Previous XVC works generally take an average speaker embedding to condition the speaker identity, which does not account for the changing timbre of speech that occurs with different pronunciations. To address this, our method uses both global and local speaker embeddings to capture the timbre changes during speech conversion. Additionally, we observed a connection between timbre and pronunciation in different languages and utilized this by incorporating a timbre encoder and a pronunciation matching network into our model. Furthermore, we found that the variation in tones is not adequately reflected in a sentence, and therefore, we used multiple references to better capture the range of a speaker’s voice. The proposed method outperformed existing systems in terms of both speech quality and speaker similarity, highlighting the effectiveness of leveraging reference information in cross-lingual voice conversion. The converted speech samples can be found on the website: http://*** © 2024, CC BY-NC-SA.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

On-Line System of Garbage Image-Orientated Intelligent Classification, Submission and Examination

On-Line System of Garbage Image-Orientated Intelligent Class...

引用

IEEE International Conference on e-Business Engineering (ICEBE)

作者： Jiayin Tian Yaozhi Wang Jiaxin Liu Yan Chen School of Computer Science and Technology Xi'an Jiaotong University Xi'an Shaanxi China Shaanxi Key Lab of Big Data Knowledge Engineering Xi'an Jiaotong University Xi'an Shaanxi China Shaanxi Key Lab of Big Data Knowledge Engineering School of Computer Science and Technology Xi'an Jiaotong University Xi'an Jiaotong University Xi'an Shaanxi China

ISBN: (数字)9798350365856

ISBN: (纸本)9798350365863

In a world brimming with new products continually, novel waste types are ubiquitous. This makes current image-based garbage classification systems difficult to perform well due to the long-tailed effects of distribution of garbage types, and necessitates an urgent and efficient garbage classification with abilities of detecting new and rare wastes and class-incremental learning for environmental sustainability. Therefore, we propose a framework of Online System of Garbage Image-Oriented Intelligent Classification, Submission, and Examination, facilitating the incremental garbage classification efforts. In which, to identify novel garbage effectively, we also introduced few-shot object detection method with two key algorithms: Two-Stage Object Detection Learning Algorithm and Dynamic Query-based Incremental Few-shot Learning Algorithm. Our experiment results show that Both outperform the current existing ones in dataset, MS COCO. Then, a strategy of Class-Incremental learning based Residual Network is proposed to meet the need of new waste class-incremental learning. The experimental results support our strategy. Finally, a prototype system employed the above algorithms and the strategy is described.

关键词： Image recognition Heavily-tailed distribution Heuristic algorithms Green products Prototypes Object detection Classification algorithms Object recognition Few shot learning Residual neural networks

来源：评论

学校读者我要写书评

暂无评论

DarkSAM: fooling segment anything model to segment nothing 24

DarkSAM: fooling segment anything model to segment nothing

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Ziqi Zhou Yufei Song Minghui Li Shengshan Hu Xianlong Wang Leo Yu Zhang Dezhong Yao Hai Jin National Engineering Research Center for Big Data Technology and System and Services Computing Technology and System Lab and Cluster and Grid Computing Lab and School of Computer Science and Technology Huazhong University of Science and Technology School of Cyber Science and Engineering Huazhong University of Science and Technology School of Software Engineering Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System and Services Computing Technology and System Lab and Hubei Engineering Research Center on Big Data Security and Hubei Key Laboratory of Distributed System Security and School of Cyber Science and Engineering Huazhong University of Science and Technology School of Information and Communication Technology Griffith University

ISBN: (纸本)9798331314385

Segment Anything Model (SAM) has recently gained much attention for its outstanding generalization to unseen data and tasks. Despite its promising prospect, the vulnerabilities of SAM, especially to universal adversarial perturbation (UAP) have not been thoroughly investigated yet. In this paper, we propose DarkSAM, the first prompt-free universal attack framework against SAM, including a semantic decoupling-based spatial attack and a texture distortion-based frequency attack. We first divide the output of SAM into foreground and background. Then, we design a shadow target strategy to obtain the semantic blueprint of the image as the attack target. DarkSAM is dedicated to fooling SAM by extracting and destroying crucial object features from images in both spatial and frequency domains. In the spatial domain, we disrupt the semantics of both the foreground and background in the image to confuse SAM. In the frequency domain, we further enhance the attack effectiveness by distorting the high-frequency components (i.e., texture information) of the image. Consequently, with a single UAP, DarkSAM renders SAM incapable of segmenting objects across diverse images with varying prompts. Experimental results on four datasets for SAM and its two variant models demonstrate the powerful attack capability and transferability of DarkSAM. Our codes are available at: https://***/CGCL-codes/DarkSAM.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Subpart Suppression Network for Few-Shot Object Counting

Subpart Suppression Network for Few-Shot Object Counting

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Lanxin Liu Xinyan Liu Guorong Li School of Computer Science and Technology Key Lab of Big Data Mining and Knowledge Management University of Chinese Academy of Sciences School of Computer Science and Technology Harbin Institute of Technology (Weihai)

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Few-shot object counting and detection aim to count objects along with their bounding boxes specified by exemplar bounding boxes. Current mainstream methods predict density maps by applying similarity between exemplar and image features to get counting results and detecting peak points from the density map as the positions of objects. However, the sub-parts of objects can also have high similarity to the exemplars, harming the counting and detection performance. To address these issues, we propose a two-stage Subpart Suppression Network (SSN), consisting of a Subpart Suppression Density map Predictor (SSDP), which enforces the model focus on whole objects rather than subparts, and a SAM-based Detector and Verifier, which refines the final prediction by a clustering method. Extensive experiments on two popular datasets show an advantage in performance over the state-of-the-art methods and prove the components’ effectiveness.

关键词： Accuracy Filtering Convolution Clustering methods Detectors Predictive models Feature extraction Acoustics Speech processing Kernel

来源：评论

学校读者我要写书评

暂无评论

Stealthy Backdoor Attack Towards Federated Automatic Speaker Verification

Stealthy Backdoor Attack Towards Federated Automatic Speaker...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Longling Zhang Lyqi Liu Dan Meng Jun Wang Shengshan Hu School of Cyber Science and Engineering Huazhong University of Science and Technology China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security OPPO Research Institute China

Automatic speech verification (ASV) authenticates individuals based on distinct vocal patterns, playing a pivotal role in many applications such as voice-based unlocking systems for devices. The ASV system comprises three stages: training, registration, and validation. The model refines using voice data in training, extracts vocal features in registration, and contrasts these with speech patterns in validation. Modern ASV models, primarily grounded in DNN architectures, require extensive data for training. Federated learning (FL) fosters model-sharing across multiple clients while ensuring data privacy. Due to its open architecture, FL is vulnerable to backdoor attacks. However, training a stealthy backdoor attack in FL presents challenges, including diminished attack generalization owing to data heterogeneity, and conspicuous triggers that render them easily detectable. In this paper, we propose a Federated Stealthy Backdoor Attack method ($FedSBA$). FedSBA aims to improve the attack model’s generalization, enhance its persistence, and elude anomaly detection under the heterogeneous data distribution. FedSBA constructs an attack model based on a personalized transformer and encompasses a stealthy trigger. Moreover, we also propose a defensive strategy that utilizes an adaptive weight aggregation scheme. The stealthiness and effectiveness of FedSBA are demonstrated by exhibiting superior performance in comparison to previous works.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Enhancing Fine-Tuning Based Backdoor Defense with Sharpness-Aware Minimization

arXiv

引用

arXiv 2023年

作者： Zhu, Mingli Wei, Shaokui Shen, Li Fan, Yanbo Wu, Baoyuan School of Data Science Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen China JD Explore Academy China Tencent AI Lab United States

Backdoor defense, which aims to detect or mitigate the effect of malicious triggers introduced by attackers, is becoming increasingly critical for machine learning security and integrity. Fine-tuning based on benign data is a natural defense to erase the backdoor effect in a backdoored model. However, recent studies show that, given limited benign data, vanilla fine-tuning has poor defense performance. In this work, we provide a deep study of fine-tuning the backdoored model from the neuron perspective and find that backdoor-related neurons fail to escape the local minimum in the fine-tuning process. Inspired by observing that the backdoor-related neurons often have larger norms, we propose FT-SAM, a novel backdoor defense paradigm that aims to shrink the norms of backdoor-related neurons by incorporating sharpness-aware minimization with fine-tuning. We demonstrate the effectiveness of our method on several benchmark datasets and network architectures, where it achieves state-of-the-art defense performance. Overall, our work provides a promising avenue for improving the robustness of machine learning models against backdoor attacks. Copyright © 2023, The Authors. All rights reserved.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

AegonKV: a high bandwidth, low tail latency, and low storage cost KV-separated LSM store with SmartSSD-based GC offloading 25

AegonKV: a high bandwidth, low tail latency, and low storage...

引用

Proceedings of the 23rd USENIX Conference on File and Storage Technologies

作者： Zhuohui Duan Hao Feng Haikun Liu Xiaofei Liao Hai Jin Bangyu Li National Engineering Research Center for Big Data Technology and System Service Computing Technology and System Lab/Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology China

ISBN: (纸本)9781939133458

The key-value separation is renowned for its significant mitigation of the write amplification inherent in traditional LSM trees. However, KV separation potentially increases performance overhead in the management of Value region, especially for garbage collection (GC) operation that is used to reduce the redundant space occupation. In response, many efforts have been made to optimize the GC mechanism for KV separation. However, our analysis indicates that such solution based on trade-offs between CPU and I/O overheads cannot simultaneously satisfy the three requirements of KV separated systems in terms of throughput, tail latency, and space usage. This limitation hinders their real-world *** this paper, we introduce AegonKV, a "three-birds-one-stone" solution that comprehensively enhances the throughput, tail latency, and space usage of KV separated systems. AegonKV first proposes a SmartSSD-based GC offloading mechanism to enable asynchronous GC operations without competing with LSM read/write for bandwidth or CPU. AegonKV leverages offload-friendly data structures and hardware/ software execution logic to address the challenges of GC offloading. Experiments demonstrate that AegonKV achieves the largest throughput improvement of 1.28-3.3 times, a significant reduction of 37%-66% in tail latency, and 15%-85% in space overhead compared to existing KV separated systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Scalable, Efficient, and Robust Dynamic Memory Management Library for HLS-based FPGAs 57

A Scalable, Efficient, and Robust Dynamic Memory Management ...

引用

57th Annual IEEE/ACM International Symposium on Microarchitecture, MICRO 2024

作者： Wang, Qinggang Zheng, Long An, Zhaozeng Xiong, Shuyi Wang, Runze Huang, Yu Yao, Pengcheng Liao, Xiaofei Jin, Hai Xue, Jingling School of Computer Science and Technology Huazhong University of Science and Technology Natl. Eng. Res. Ctr. for Big Data Technol. and Syst./Serv. Comp. Technol. and Syst. Lab/Cluster and Grid Comp. Lab Wuhan430074 China Zhejiang Lab Hangzhou311121 China School of Computer Science and Engineering University of New South Wales SydneyNSW2052 Australia

ISBN: (纸本)9798350350579

Nowadays, high-level synthesis (HLS) has gained prominence for FPGA-based architecture prototyping, enhancing productivity significantly. Despite this advancement, HLS tools are impeded by a critical drawback: They lack support for dynamic memory management (DMM), leading to static mem-ory allocation and suboptimal use of memory resources. In response, numerous efforts have been made to develop DMM solutions compatible with HLS. However, our analysis indicates that existing solutions fail to concurrently meet the desired trifecta of scalability (efficient management of memory of any size), efficiency (minimal latency in memory (de-)allocation), and robustness (low allocation failure rates). This limitation hampers their applicability in real-world scenarios. In this paper, we introduce GraDMM, a 'three-birds-one-stone' solution that comprehensively enhances the scalability, efficiency, and robustness of DMM. The key insight is to formulate memory (de-)allocation as graph analytics and lever-Age sophisticated FPGA-based graph processing techniques. To achieve scalability, GraDMM specializes a simplified pipeline that significantly suppresses resource utilization expansion caused by managed memory scaling. This is crucial for managing arbitrarily sized memory on resource-limited FPGA platforms. For efficiency, GraDMM implements a data-centric concurrent traversal scheme and a shortcut-Assisted fast traversal policy to accelerate (de-)allocation-guided graph traversal, reducing mem-ory (de-)allocation latency. To enhance robustness, GraDMM incorporates an adaptive memory defragmenter that defragments managed memory to minimize fragmentation-induced allocation failures. GraDMM is encapsulated as a library, providing high-level interfaces for users and ensuring synthesizability with Vi-vado HLS. Experimental results demonstrate that GraDMM out-performs three state-of-The-Art HLS-compatible DMM solutions by significant margins: 56.71 %-85.59% in resource consumption savin

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Boosting Backdoor Attack with A Learnable Poisoning Sample Selection Strategy

arXiv

引用

arXiv 2023年

作者： Zhu, Zihao Zhang, Mingda Wei, Shaokui Shen, Li Fan, Yanbo Wu, Baoyuan School of Data Science Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen China JD Explore Academy China Tencent AI Lab United States

data-poisoning based backdoor attacks aim to insert backdoor into models by manipulating training datasets without controlling the training process of the target model. Existing attack methods mainly focus on designing triggers or fusion strategies between triggers and benign samples. However, they often randomly select samples to be poisoned, disregarding the varying importance of each poisoning sample in terms of backdoor injection. A recent selection strategy filters a fixed-size poisoning sample pool by recording forgetting events, but it fails to consider the remaining samples outside the pool from a global perspective. Moreover, computing forgetting events requires significant additional computing resources. Therefore, how to efficiently and effectively select poisoning samples from the entire dataset is an urgent problem in backdoor attacks. To address it, firstly, we introduce a poisoning mask into the regular backdoor training loss. We suppose that a backdoored model training with hard poisoning samples has a more backdoor effect on easy ones, which can be implemented by hindering the normal training process (i.e., maximizing loss w.r.t. mask). To further integrate it with normal training process, we then propose a learnable poisoning sample selection strategy to learn the mask together with the model parameters through a min-max optimization. Specifically, the outer loop aims to achieve the backdoor attack goal by minimizing the loss based on the selected samples, while the inner loop selects hard poisoning samples that impede this goal by maximizing the loss. After several rounds of adversarial training, we finally select effective poisoning samples with high contribution. Extensive experiments on benchmark datasets demonstrate the effectiveness and efficiency of our approach in boosting backdoor attack performance. © 2023, CC BY.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

RTUNet++: Assessment of Osteosarcoma MRI Image Segmentation leveraging Hybrid CNN-Transformer Approach with Dense Skip Connection

RTUNet++: Assessment of Osteosarcoma MRI Image Segmentation ...

引用

IEEE International Conference on Signal and Image Processing (ICSIP)

作者： Binfeng Zou Yifei Chen Zhanghao Chen Yifei Sun Yifan Huang Feiwei Qin Changmiao Wang HDU-ITMO Joint Institute Hangzhou Dianzi University Hangzhou China School of Computer Science and Technology Hangzhou Dianzi University Hangzhou China Medical Big Data Lab Shenzhen Research Institute of Big Data Shenzhen China

Orthopedic osteosarcoma is a prevalent malignant bone tumor. Preoperative planning, efficacy evaluation, and metastasis detection of osteosarcoma necessitate the use of magnetic resonance imaging (MRI). Due to the varying 10-cations, structures, sizes, and shapes of osteosarcomas among patients, as well as the high degree of tumor heterogeneity, segmenting osteosarcoma images manually presents a significant challenge for clinicians. In the meantime, the grayscale and texture features within the tumor in MRI images are not homogeneous, resulting in small grayscale differences between the tumor tissue and the surrounding normal tissue, which also presents a significant challenge for deep learning-based image segmentation methods. This paper proposes RTUNet++, a novel method for segmenting osteosarcoma MRI images. This method integrates the ResNet residual module, Transformer attention mechanism, and UNet++ Dense Skip Connection structure. We employ a CNN-Transformer hybrid architecture to prevent the loss of spatial feature data. Features with different semantic scales are aggregated at the decoder using multi-fusion dense jump connections to achieve flexible feature fusion. Consequently, in comparison experiments with various classical models, our proposed RTUNet++ achieves superior segmentation results on an osteosarcoma MRI image dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：