In this work, we propose a no-reference image quality evaluation approach, aiming to solve the problem that the traditional convolutional neural network is insufficient to express the global information of the image. ...
详细信息
The Area Under the ROC Curve (AUC) is a well-known metric for evaluating instance-level long-tail learning problems. In the past two decades, many AUC optimization methods have been proposed to improve model performan...
ISBN:
(纸本)9798331314385
The Area Under the ROC Curve (AUC) is a well-known metric for evaluating instance-level long-tail learning problems. In the past two decades, many AUC optimization methods have been proposed to improve model performance under long-tail distributions. In this paper, we explore AUC optimization methods in the context of pixel-level long-tail semantic segmentation, a much more complicated scenario. This task introduces two major challenges for AUC optimization techniques. On one hand, AUC optimization in a pixel-level task involves complex coupling across loss terms, with structured inner-image and pairwise inter-image dependencies, complicating theoretical analysis. On the other hand, we find that mini-batch estimation of AUC loss in this case requires a larger batch size, resulting in an unaffordable space complexity. To address these issues, we develop a pixel-level AUC loss function and conduct a dependency-graph-based theoretical analysis of the algorithm's generalization ability. Additionally, we design a Tail-Classes Memory Bank (T-Memory Bank) to manage the significant memory demand. Finally, comprehensive experiments across various benchmarks confirm the effectiveness of our proposed AUCSeg method. The code is available at https://***/boyuh/AUCSeg.
As the material basis for all life activities, proteins play a crucial role in performing life activities. Most real-life proteins perform various functions in the form of protein complexes, so it is essential for und...
As the material basis for all life activities, proteins play a crucial role in performing life activities. Most real-life proteins perform various functions in the form of protein complexes, so it is essential for understanding of life activities to accurately identify protein complexes. Most of the existing approaches learn protein interactions directly from protein topology,ignoring the biological characteristics of the actual proteins. Nevertheless, actual protein complexes are usually composed of protein cores and protein attachments. In this study, a new MP-DE algorithm is designed from the core-attachment structure to generate protein complex cores using Markov clustering, while searching for attached proteins in the second-order neighborhood of core protein complexes based on differential evolution(DE) algorithm. The experimental results show that the proposed method has high detection accuracy and efficiency on the current mainstream protein-protein interaction(PPI) database.
Reconfigurable intelligent surface (RIS) is an emerging technology for improving the performance of wireless communications. In this letter, a joint Squared-Sine function and atomic norm minimization (ANM), denoted as...
详细信息
In the paper, we firstly study the algebraic structures of ppk-additive cyclic codes and give the generator polynomials and the minimal spanning set of these codes. Secondly, a necessary and sufficient condition for t...
详细信息
Large language model (LLM) has achieved promising performance in multilingual machine translation tasks through zero/few-shot prompts or prompt-tuning. However, due to the mixture of multilingual data during the pre-t...
详细信息
A $\boldsymbol{40}\times\boldsymbol{40}\times \boldsymbol{1.6}\mathbf{mm}^{3}$ ultra-wideband MIMO antenna is proposed in this paper. The antenna realizes four-notched band characteristics by embedding different typ...
详细信息
A $\boldsymbol{40}\times\boldsymbol{40}\times \boldsymbol{1.6}\mathbf{mm}^{3}$ ultra-wideband MIMO antenna is proposed in this paper. The antenna realizes four-notched band characteristics by embedding different types of slits in the radiating element. The antenna has good UWB performance, with four-band rejection to avoid interference caused by four narrow frequency bands WiMAX(3.3-3.8GHz), WLAN(S.1-S.8GHz), INSAT(6.7-7.3GHz) and ITV 8GHz (8.13-8.38GHz). It achieves better than - 15dB of isolation, and impedance bandwidth from 3.06-15.5GHz.
The identification of drug-protein interactions(DTIs) is a critical step in drug development and ***,detecting these interactions using scientific methods presents a formidable *** models often struggle to extract rel...
The identification of drug-protein interactions(DTIs) is a critical step in drug development and ***,detecting these interactions using scientific methods presents a formidable *** models often struggle to extract relevant information from drug-protein sequences,and they fail to consider pre-existing interactions between ***,we introduce a deep learning model,DPAG,based on attention mechanisms and graph neural networks,which overcomes these *** leverages information from both the drug and protein sequences,as well as the association between drug-protein pairs,to predict *** conducted multiple experiments to evaluate the performance of the model in benchmark datasets,and the results showed that DPAG achieved excellent performance in terms of AUC,AUPR,precision,and *** findings suggest that DPAG holds significant potential as a promising tool for predicting DTIs in drug development and *** is available at https://***/veres se/DPAG.
In this paper, several conjectures proposed in [2] are studied, involving the equivalence and duality of polycyclic codes associated with trinomials. According to the results, we give methods to construct isodual and ...
详细信息
To address the significant performance degradation of conventional underdetermined blind source separation algorithms for frequency-hopping (FH) signals under time-frequency (TF) overlapping conditions, this paper pre...
详细信息
暂无评论