1,721,060 research outputs found
Watermarking Protocol for Deep Neural Network Ownership Regulation in Federated Learning
No full textWith the wide application of deep learning models, it is important to verify an author's possession over a deep neural network model by watermarks and protect the model. The development of distributed learning paradigms such as federated learning raises new challenges for model protection. Each author should be able to conduct independent verification and trace traitors. To meet those requirements, we propose a watermarking protocol, Merkle-Sign to meet the prerequisites for ownership verification in federated learning. Our work paves the way for generalizing watermark as a practical security mechanism for protecting deep learning models in distributed learning platforms.No Full Tex
Message from the General Chairs
No full textPresents the introductory welcome message from the conference proceedings. May include the conference officers' congratulations to all involved with the conference event and publication of the proceedings record.No Full Tex
Graphics processing unit acceleration of the island model genetic algorithm using the CUDA programming platform
No full textGenetic algorithms are a practical approach for finding near-optimal solutions for nondeterministic polynomial-hard problems. In this work we exploit the parallel processing capability of graphics processing units and Nvidia's CUDA programming platform to accelerate the island model genetic algorithm by modifying the evolutionary operations to fit the hardware architecture and have successfully achieved significant computational speedups.No Full Tex
Lip Image Segmentation Based on a Fuzzy Convolutional Neural Network
No full textResearch has shown that the human lip and its movements are a rich source of information related to speech content and speaker's identity. Lip image segmentation, as a fundamental step in many lip-reading and visual speaker authentication systems, is of vital importance. Because of variations in lip color, lighting conditions and especially the complex appearance of an open mouth, accurate lip region segmentation is still a challenging task. To address this problem, this article proposes a new fuzzy deep neural network having an architecture that integrates fuzzy units and traditional convolutional units. The convolutional units are used to extract discriminative features at different scales to provide comprehensive information for pixel-level lip segmentation. The fuzzy logic modules are employed to handle various kinds of uncertainties and to provide a more robust segmentation result. An end-to-end training scheme is then used to learn the optimal parameters for both the fuzzy and the convolutional units. A dataset containing more than 48 000 images of various speakers, under different lighting conditions, was used to evaluate lip segmentation performance. According to the experimental results, the proposed method achieves state-of-the-art performance when compared with other algorithms.Full Tex
Lip Image Segmentation in Mobile Devices Based on Alternative Knowledge Distillation
No full textLip image segmentation, as the first step in many lip-related tasks (e.g. automatic lipreading), is of vital significance for the subsequent procedures. Nowadays, with the increasing computational power of the mobile devices, mobile applications become more and more popular. In this paper, a new approach is proposed, which is able to segment the lip region in natural scenes and is of acceptable computational complexity to be implemented in mobile devices. Two networks including a complex teacher network and a compact student network with the same structure are employed. With the proposed remedy loss and the alternative knowledge distillation scheme, the student network can learn useful knowledge from the teacher network effectively and efficiently, and even rectify some of its segmentation errors. A dataset containing 49 people captured under natural scenes by various cellphone cameras is adopted for evaluation and the experiment results have demonstrated that the proposed student network even outperforms the teacher network with much less computational cost.No Full Tex
Current trends of granular data mining for biomedical data analysis
Full text linkBiomedical data are available in many different formats, including numeric, textual reports, signals or images, and they come available from a variety of sources. Biomedical data typically suffer from incompleteness, uncertainty and vagueness, posing several challenges to perform data analysis, such high dimensionality, class imbalance or low numbers of samples [ 1 , 2 ]. Granular Computing, the term coined by Prof. L. A. Zadeh, provides a powerful tool for multiple granularity and multiple-view data analysis, which is of vital importance for understanding data driven analysis at different levels of ab- straction (granularity) [3] . It is worth stressing that human’s capabilities in effective information or ganization and efficient reasoning with complex and uncertain information is highly dependent on hierarchical Granular Computing [ 4 , 5 ]. We have been witnessing significant advances of Granular Computing in the scientific and engineering domains. Data mining based on Granular Computing in biomedical data analysis is an emerging field which crosses multiple research disciplines and in- dustry domains. As a meta-mathematical methodology, granular data mining provides a theoretical framework for biomed- ical data analytics. It helps to extract knowledge when we are provided with an insufficient data that may also contain a significant amount of unstructured, uncertain and imprecise data. Granular data mining technology has exhibited some strong capabilities and advantages in intelligent data analysis and uncertainty reasoning for biomedical data. However, de- termining how to integrate Granular Computing and data mining to combine their advantages remains an interesting and important research topic. Recent survey indicated that granular data mining research has been focused on exploring the advantages, and also the challenges, derived from collecting and mining vast amounts of available biomedical data sources. It has therefore become strongly and timely justified to develop theoretical models and practical algorithms for carrying out granular data mining for biomedical data analysis.Full Tex
Leveraging Multi-task Learning for Unambiguous and Flexible Deep Neural Network Watermarking
No full textDeep neural networks are playing an important role in many real-life applications. An important prerequisite in commercializing deep neural networks is the identification of their genuine owners. Therefore, watermarking schemes that embed the owner's identity information into the models have been proposed. However, current schemes cannot meet all the security requirements such as unambiguity and are inflexible since most of them focus on classification models. To meet the formal definitions of the security requirements and increase the applicability of deep neural network watermarking schemes, we propose a new method, MTLSign, based on multi-task learning. By treating the watermark embedding as an extra task, the security requirements are explicitly formulated and met with well-designed regularizers and components from cryptography. Experiments have demonstrated that MTLSign is flexible and robust for practical security in machine learning applications.Full Tex
Feature Extraction for Visual Speaker Authentication Against Computer-Generated Video Attacks
No full textRecent research shows that the lip feature can achieve reliable authentication performance with a good liveness detection ability. However, with the development of sophisticated human face generation methods by the deepfake technology, the talking videos can be forged with high quality and the static lip information is not reliable in such case. Meeting with such challenge, in this paper, we propose a new deep neural network structure to extract robust lip features against human and Computer-Generated (CG) imposters. Two novel network units, i.e. the feature-level Difference block (Diffblock) and the pixel-level Dynamic Response block (DRblock), are proposed to reduce the influence of the static lip information and to represent the dynamic talking habit information. Experiments on the GRID dataset have demonstrated that the proposed network can extract discriminative and robust lip features and outperform two state-of-the-art visual speaker authentication approaches in both human imposter and CG imposter scenarios.No Full Tex
Fine-Grained Lip Image Segmentation using Fuzzy Logic and Graph Reasoning
No full textFine-grained lip image segmentation plays a critical role in downstream tasks such as automatic lipreading, as it enables the accurate identification of inner mouth components such as teeth and tongue which are essential for comprehending spoken utterances. However, achieving accurate and robust lip image segmentation in natural scenes is still challenging due to significant variations in lighting condition, head pose and background. This paper proposes a novel deep neural network based method for fine-grained lip image segmentation that exploits fuzzy and graph theories to handle these variations. A fuzzy learning module is designed to deal with the uncertainties in color and edge information and enhance feature maps at various scales. The fuzzy graph reasoning module with fuzzy projection models the relationship among semantics components and achieves a global receptive field. In our experiments, a fine-grained lip region segmentation dataset, i.e., FLRSeg, is built for evaluation and experiment results have shown that the proposed method can achieve superior segmentation performance (94.36% in pixel accuracy and 74.89% in mIoU) compared with several SOTA lip image segmentation methods.Full Tex
A novel online ensemble convolutional neural networks for streaming data
No full textIn this study, we introduce an online ensemble method based on convolutional neural networks (CNNs) for streaming data. Recent work has shown that a convolution operation has been an effective way to extract features. In particular, we proposed a CNN working in an online manner as a base classifier. Then, an ensemble approach is devised to boost the performance of all base classifiers. We also propose two loss terms which can adapt to the imbalanced data stream as well as handling the forgetting issue of deep networks. The experiments conducted on a number of datasets chosen from different sources demonstrate that the proposed ensemble approach performs significantly better than a single network and some well-known online learning algorithms including additive models and Online Bagging.No Full Tex
- …
