Federated linear regressions have been developed and applied in various domains, where multiparties collaboratively and securely perform optimization algorithms, e.g., Gradient Descent, to learn a set of optimal model...
详细信息
Pose estimation is important for robotic perception, path planning, etc. Robot poses can be modeled on matrix Lie groups and are usually estimated via filter-based methods. In this paper, we establish the closed-form ...
详细信息
In this paper, we study the statistical properties of distributed kernel ridge regression together with random features (DKRR-RF), and obtain optimal generalization bounds under the basic setting, which can substantia...
详细信息
Currently, screen content video applications are widely used in our daily lives. As the latest Screen Content Coding (SCC) standard, Versatile Video Coding (VVC) SCC employs a quad-tree plus nested multi-type tree (QT...
详细信息
Currently, screen content video applications are widely used in our daily lives. As the latest Screen Content Coding (SCC) standard, Versatile Video Coding (VVC) SCC employs a quad-tree plus nested multi-type tree (QTMT) coding structure and various screen content coding modes (CMs). This design enhances the coding efficiency of VVC SCC but also results in a highly complex coding process, which significantly hinders the broader adoption of screen content video technology. Consequently, improving the coding speed of VVC SCC is highly desirable. In this paper, we propose a fast CM and transform decision algorithm for Intra prediction in VVC SCC. Specifically, we initially use Convolutional Neural Networks (CNNs) to predict content types for all Coding Units (CUs). Subsequently, we predict candidate CMs for CUs based on the CM distributions of different content types. We then select the Sum of Absolute Transformed Difference (SATD) as a feature and use a naive Bayes classifier to skip unlikely Intra mode early. Finally, we terminate Block-based Differential Pulse-Code Modulation (BDPCM) early and then select the best transform type in Intra mode prediction to improve coding speed. Experimental results demonstrate that the proposed algorithm improves coding speed by an average of 39.28%, with the BDBR increasing by 0.80%.
The miniaturization and integration of beam steering devices have consistently been the focus of the field. Conventional methods alter the eigenmode of the optical cavity by regulating the refractive index. Due to the...
详细信息
Interference and scattering, often deemed undesirable, are inevitable in wireless communications, especially when the current mobile networks and upcoming sixth generation (6G) have turned into ultra-dense networks. C...
详细信息
Communication efficiency is one of the key bottlenecks in Federated Learning (FL). Compression techniques, such as sparsification and quantization, are used to reduce communication overhead. However, joint designs of ...
Communication efficiency is one of the key bottlenecks in Federated Learning (FL). Compression techniques, such as sparsification and quantization, are used to reduce communication overhead. However, joint designs of these techniques under communication constraints are not well-explored. This paper investigates the joint uplink compression problem in communication-constrained FL systems. We propose a Block-TopK sparsification scheme to reduce the proportion of bits used for locating entries of a sparsified vector. Considering the communication constraints, an optimization formulation is proposed to minimize the compression error. By solving the optimization problem, our joint compression method provides a better trade-off between sparsity budget and bit width. Numerical results demonstrate that our approach achieves 99.96% of baseline accuracy with only 1.56% of the baseline communication overhead when training ResNet-18 on CIFAR-10.
Channel modeling is indispensable in a communication system. In this paper, a novel scheme for channel modeling using quantum generative adversarial model was proposed. A quantum generative adversarial network is a ge...
详细信息
ISBN:
(纸本)9781665450867
Channel modeling is indispensable in a communication system. In this paper, a novel scheme for channel modeling using quantum generative adversarial model was proposed. A quantum generative adversarial network is a generative adversarial model with a quantum circuit as the generative module and a deep neural network as the discriminant module, thereby exploiting the privilege of quantum algorithms in simulating probability distributions to stochastic channel models. Experiments were conducted on IBM QX quantum computing platform. The gradient descent of the cost function and Kullback-Leibler divergence were analyzed. Results verify the feasibility and superiority of the quantum generative adversarial network for channel modeling.
Tumor localization and lymph node metastasis (LNM) diagnosis are two important tasks for gynecologist to make decisions in cervical cancer treatments. Aiming to develop an accurate and convenient diagnosis system, we ...
Tumor localization and lymph node metastasis (LNM) diagnosis are two important tasks for gynecologist to make decisions in cervical cancer treatments. Aiming to develop an accurate and convenient diagnosis system, we propose a multi-task residual cross-attention network named MRCNet for tumor segmentation and LNM prediction. Specifically, we tackle task correlation with underlying related supervision information, and capture multi-level features by multi-scale convolutional neural network, which equipped with cross-attention module concerning spatial and channel dimensions to emphasize meaningful features. A total of 1123 cervical cancer patients from 13 centers in China are collected to assess the architecture, 2 centers of them were set as an external testing cohort. The experimental results demonstrate the promising inference performance and generalization ability of our MRCNet for both segmentation and classification tasks, which can help doctors make judgments about treatment measures.
暂无评论