In the field of heterogeneous federated learning (FL), the key challenge is to efficiently and collaboratively train models across multiple clients with different data distributions, model structures, task objectives,...
详细信息
Commit messages are important complementary information used in understanding code changes. To address message scarcity, some work is proposed for automatically generating commit messages. However, most of these appro...
详细信息
Commit messages are important complementary information used in understanding code changes. To address message scarcity, some work is proposed for automatically generating commit messages. However, most of these approaches focus on generating summary of the changed software entities at the superficial level, without considering the intent behind the code changes (e.g., the existing approaches cannot generate such message:"fixing 'null' pointer exception"). Considering developers often describe the intent behind the code change when writing the messages, we propose ChangeDoc, an approach to reuse existing messages in version control systems for automatical commit message generation. Our approach includes syntax, semantic, pre-syntax, and pre-semantic similarities. For a given commit without messages, it is able to discover its most similar past commit from a large commit repository, and recommend its message as the message of the given commit. Our repository contains half a million commits that were collected from SourceForge. We evaluate our approach on the commits from 10 projects. The results show that 21.5% of the recommended messages by ChangeDoc can be directly used without modification, and 62.8% require minor modifications. In order to evaluate the quality of the commit messages recommended by ChangeDoc, we performed two empirical studies involving a total of 40 participants (10 professional developers and 30 students). The results indicate that the recommended messages are very good approximations of the ones written by developers and often include important intent information that is not included in the messages generated by other tools.
With the advancement of image authoring software, the traces of image tampering operation have become increasingly difficult to detect. To enhance the performance of manipulation detection, we propose a novel end-to-e...
详细信息
The propagation of spaceborne synthetic aperture radar (SAR) signal is affected by the ionosphere, which will lead to the degradation of the imaging quality, especially in the conditions of low-frequency and wide-band...
详细信息
Due to the heterogeneous distributions of multi-center rs-fMRI data, currently it is still challenging to identify accurately autism spectrum disorder (ASD) patients from these heterogenous data. To deal with the infl...
详细信息
作者：
Su, CanXue, XinleiMa, LeiZhang, XiaolongYan, WeiBian, KaiguiPeking University
School of Computer Science AI Innovation Center National Engineering Laboratory for Big Data Analysis and Applications Beijing100871 China Peking University
School of Computer Science Beijing100871 China Peking University
Beijing Academy of Artificial Intelligence National Biomedical Imaging Center College of Future Technology National Key Laboratory for Multimedia Information Processing Beijing100871 China Beihang University
Beijing100191 China
Existing person re-identification (ReID) methods mainly rely on images and videos to match persons across cameras, yet visual data captured by cameras are vulnerable to environmental interferences (e.g. illumination a...
详细信息
Alzheimer’s disease is a progressive form of memory loss that worsens over time. Detecting it early, when memory issues are mild, is crucial for effective interventions. Recent advancements in computer technology, sp...
详细信息
Federated learning (FL) is an emerging machine learning paradigm for preserving data privacy. However, diverse client hardware often has varying computation resources. Such system heterogeneity limits the participatio...
Federated learning (FL) is an emerging machine learning paradigm for preserving data privacy. However, diverse client hardware often has varying computation resources. Such system heterogeneity limits the participation of resource-constrained clients in FL, and hence degrades the global model accuracy. To enable heterogeneous clients to participate in and contribute to FL training, previous works tackle this problem by assigning customized sub-models to individual clients with model pruning, distillation, or low-rank based techniques. Unfortunately, the global model trained by these methods still encounters performance degradation due to heterogeneous sub-model aggregation. Besides, most methods are heuristic-based and lack convergence analysis. In this work, we propose the FedLMT framework to bridge the performance gap, by assigning clients with a homogeneous pre-factorized low-rank model to substantially reduce resource consumption without conducting heterogeneous aggregation. We theoretically prove that the convergence of the low-rank model can guarantee the convergence of the original full model. To further meet clients' personalized resource needs, we extend FedLMT to pFedLMT, by separating model parameters into common and custom ones. Finally, extensive experiments are conducted to verify our theoretical analysis and show that FedLMT and pFedLMT outperform other baselines with much less communication and computation costs.
Narrow-band interference (NBI) is a common form of interference in SAR systems, severely reducing the quality of SAR images. The NBI suppression method based on empirical mode decomposition (EMD) can adaptively split ...
详细信息
ISBN:
(纸本)9781665468893
Narrow-band interference (NBI) is a common form of interference in SAR systems, severely reducing the quality of SAR images. The NBI suppression method based on empirical mode decomposition (EMD) can adaptively split the raw signal into multiple independent intrinsic modal functions (IMFs), and reconstruct the useful signal after removing the IMFs of the NBI. However, the mode mixing problem it brings will cause part of the useful signal to be removed simultaneously when removing the NBI. A novel NBI suppression method based on complementary ensemble EMD (CEEMD) is proposed in this paper to solve this problem. First, the moment kurtosis coefficient is used to detect the presence of NBI in the raw echo, and CEEMD is used to decompose the NBI-contained raw echo into a series of IMFs. Then, the second-order difference of permutation entropy (PE) of all IMFs is calculated to obtain the global threshold for distinguishing IMFs with NBI and useful signal. Finally, the target IMF is used to reconstruct useful signals to remove the NBI. Experimental results based on real C-band spaceborne SAR data with simulated NBI verify the performance of the proposed method.
data mining technology has yielded fruitful results in the area of crime discovery and intelligent decision making. Credit card is one of the most popular payment methods, providing great convenience and efficiency. H...
详细信息
暂无评论