Search

Your search keyword '"Man, Dapeng"' showing total 120 results
Start Over Author "Man, Dapeng"
120 results on '"Man, Dapeng"'

1. Revisiting Interpolation Augmentation for Speech-to-Text Generation

Author

Xu, Chen, Wang, Jie, Liu, Xiaoqian, Dong, Qianqian, Zhang, Chunliang, Xiao, Tong, Zhu, Jingbo, Man, Dapeng, and Yang, Wu

Subjects
Computer Science - Computation and Language, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

Speech-to-text (S2T) generation systems frequently face challenges in low-resource scenarios, primarily due to the lack of extensive labeled datasets. One emerging solution is constructing virtual training samples by interpolating inputs and labels, which has notably enhanced system generalization in other domains. Despite its potential, this technique's application in S2T tasks has remained under-explored. In this paper, we delve into the utility of interpolation augmentation, guided by several pivotal questions. Our findings reveal that employing an appropriate strategy in interpolation augmentation significantly enhances performance across diverse tasks, architectures, and data scales, offering a promising avenue for more robust S2T systems in resource-constrained settings., Comment: ACL 2024 Findings

Published
2024

2. Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning

Author

Liu, Tao, Zhang, Yuhang, Feng, Zhu, Yang, Zhiqin, Xu, Chen, Man, Dapeng, and Yang, Wu

Subjects
Computer Science - Cryptography and Security, Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Backdoors on federated learning will be diluted by subsequent benign updates. This is reflected in the significant reduction of attack success rate as iterations increase, ultimately failing. We use a new metric to quantify the degree of this weakened backdoor effect, called attack persistence. Given that research to improve this performance has not been widely noted,we propose a Full Combination Backdoor Attack (FCBA) method. It aggregates more combined trigger information for a more complete backdoor pattern in the global model. Trained backdoored global model is more resilient to benign updates, leading to a higher attack success rate on the test set. We test on three datasets and evaluate with two models across various settings. FCBA's persistence outperforms SOTA federated learning backdoor attacks. On GTSRB, postattack 120 rounds, our attack success rate rose over 50% from baseline. The core code of our method is available at https://github.com/PhD-TaoLiu/FCBA.

Published
2024

3. Bridging the Gaps of Both Modality and Language: Synchronous Bilingual CTC for Speech Translation and Speech Recognition

Author

Xu, Chen, Liu, Xiaoqian, He, Erfeng, Zhang, Yuhao, Dong, Qianqian, Xiao, Tong, Zhu, Jingbo, Man, Dapeng, and Yang, Wu

Subjects
Computer Science - Computation and Language, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

In this study, we present synchronous bilingual Connectionist Temporal Classification (CTC), an innovative framework that leverages dual CTC to bridge the gaps of both modality and language in the speech translation (ST) task. Utilizing transcript and translation as concurrent objectives for CTC, our model bridges the gap between audio and text as well as between source and target languages. Building upon the recent advances in CTC application, we develop an enhanced variant, BiL-CTC+, that establishes new state-of-the-art performances on the MuST-C ST benchmarks under resource-constrained scenarios. Intriguingly, our method also yields significant improvements in speech recognition performance, revealing the effect of cross-lingual learning on transcription and demonstrating its broad applicability. The source code is available at https://github.com/xuchennlp/S2T., Comment: Submitted to ICASSP 2024

Published
2023

4. FedGG: Leveraging Generative Adversarial Networks and Gradient Smoothing for Privacy Protection in Federated Learning

Author

Lv, Jiguang, Xu, Shuchun, Zhan, Xiaodong, Liu, Tao, Man, Dapeng, Yang, Wu, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Carretero, Jesus, editor, Shende, Sameer, editor, Garcia-Blas, Javier, editor, Brandic, Ivona, editor, Olcoz, Katzalin, editor, and Schreiber, Martin, editor

Published
2024
Full Text
View/download PDF

9. Identification method for malicious traffic in industrial Internet under new unknown attack scenarios.

Author

ZENG Fanyi, MAN Dapeng, XU Chen, HAN Shuai, WANG Huanran, ZHOU Xue, LI Xinchun, and YANG Wu

Abstract

Aiming at the problem of traffic data distribution shift caused by new unknown attacks in the industrial Internet, a malicious traffic identification method based on neighborhood filtering and stable learning was proposed to enhance the effectiveness and robustness of the existing graph neural network model in identifying known malicious traffic. Firstly, the graph structure of the traffic data was modeled to capture the topological relationship and interaction mode in communication behavior. Secondly, the traffic subgraph was divided based on the neighborhood filtering mechanism of biased sampling to eliminate the pseudo-homogeneity between communication behaviors. Finally, the statistical independence of high-dimensional traffic features was realized by applying graph representation learning and stable learning strategies, combined with adaptive sample weighting and collaborative loss optimization methods. The experimental results on two benchmark datasets show that compared with the baseline method, the recognition performance of the proposed method is increased by more than 2.7% in the new unknown attack scenario, which shows its high efficiency and practicability in the industrial Internet environment. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

12. A Dynamic Event Region Tracking Approach Based on Node Calibration

Author

Yu, Xiang, Lu, Hui, Wang, Le, Man, Dapeng, Hutchison, David, Editorial Board Member, Kanade, Takeo, Editorial Board Member, Kittler, Josef, Editorial Board Member, Kleinberg, Jon M., Editorial Board Member, Mattern, Friedemann, Editorial Board Member, Mitchell, John C., Editorial Board Member, Naor, Moni, Editorial Board Member, Pandu Rangan, C., Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Terzopoulos, Demetri, Editorial Board Member, Tygar, Doug, Editorial Board Member, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Sun, Xingming, editor, Pan, Zhaoqing, editor, and Bertino, Elisa, editor

Published
2019
Full Text
View/download PDF

19. A Webpage Deletion Algorithm Based on Hierarchical Filtering

Author

Chen, Xunxun, Wang, Wei, Man, Dapeng, Xuan, Sichang, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Wang, Fu Lee, editor, Lei, Jingsheng, editor, Gong, Zhiguo, editor, and Luo, Xiangfeng, editor

Published
2012
Full Text
View/download PDF

33. ECBCM: A prestige‐based edge computing blockchain security consensus model.

Author

Xuan, Shichang, Chen, Zhiyu, Chung, Ilyong, Tan, Haowen, Man, Dapeng, Du, Xiaojiang, Yang, Wu, and Guizani, Mohsen

Subjects
EDGE computing, PROBLEM solving, INFORMATION sharing, ELECTRONIC data processing, BLOCKCHAINS
Abstract

The explosive growth of data in the network has brought huge burdens and challenges to traditional centralized cloud computing data processing. To solve this problem, edge computing technology came into being. Because the edge is closer to the user, processing part of the data at the edge can also bring a faster response to the user and improve their experience. However, the existing edge computing platforms have problems such as data storage security and multiparty data mutual trust. Blockchain technology has become an important means to solve the above data storage and sharing problems due to its excellent characteristics. The core of blockchain technology is consensus, and its speed and security will directly affect the efficiency and stability of the blockchain system. Therefore, this study uses the consensus mechanism as an entry point to reduce the resource consumption of the edge computing blockchain system and improve its security. In order to reduce the resource consumption of traditional consensus algorithms, improve their adaptability in the edge computing environment, and solve the security problem caused by the concentration of node rights, a prestige‐based edge computing blockchain security consensus model (ECBCM) is proposed. ECBCM is a general model based on prestige rewards and penalties. It also introduces a node replacement mechanism to ensure the fault tolerance of the consensus process. According to the results of multiple sets of performance comparison experiments and security verification experiments after embedding the existing consensus algorithm, the validity of the consensus model is confirmed. [ABSTRACT FROM AUTHOR]

Published
2021
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results