The graph coloring problem (GCP) is a classic combinatorial optimization problem that has been widely applied in various fields such as mathematics, computer science, and biological science. Due to the NP hard nature ...
详细信息
Transformer-based real-time semantic segmentation algorithms have demonstrated significant potential. Nonetheless, current mainstream Transformer methods typically overlook the correlation between each region in the a...
Transformer-based real-time semantic segmentation algorithms have demonstrated significant potential. Nonetheless, current mainstream Transformer methods typically overlook the correlation between each region in the attention calculation process. Limiting attention to the most important regions further reduces model parameters and computational load. To address this issue, a dynamic sparse axial rectangular attention Transformer was developed for real-time semantic segmentation (DSARFormer). DSARFormer comprises two key modules, namely the DSARFormer Block and the CNN-Transformer feature fusion module (CTFM). The DSARFormer Block contains dynamic sparse axial rectangular attention (ARAttention), which calculates the attention of the most relevant rectangular regions in the horizontal and vertical directions. Meanwhile, CTFM can effectively integrate the features of CNN and Transformer, making it suitable for real-time semantic segmentation. Both modules were evaluated on the ADE20K and Cityscapes datasets. The results revealed that DSARFormer achieved 39.3% mIoU and 73.4% mIoU at 48.5FPS and 46.3FPS, respectively, outperforming current mainstream real-time semantic segmentation algorithms. Code is available at https://***/Panyw1011/DSARFormer .
RNA-binding proteins (RBPs) are essential for gene expression, and the complex RNA-protein interaction mechanisms require analysis of global RNA information. Therefore, accurate prediction of RBP binding sites on full...
详细信息
As a universal language, English has been paid more and more attention, among which oral English learning is very important. In this paper, the two key technologies of pronunciation error detection and quality evaluat...
详细信息
The significant progress of vision-language pre-trained models (VLMs) and large language models (LLMs) have provided a feasible new mode for image captioning, which relies on VLMs to process images and then utilizes L...
详细信息
Vehicle routing problem with time windows(VRPTW)is a core combinatorial optimization problem in distribution *** electric vehicle routing problem with time windows under demand uncertainty and weight-related energy co...
详细信息
Vehicle routing problem with time windows(VRPTW)is a core combinatorial optimization problem in distribution *** electric vehicle routing problem with time windows under demand uncertainty and weight-related energy consumption is an extension of the *** some researchers have studied either the electric VRPTW with nonlinear energy consumption model or the impact of the uncertain customer demand on the conventional vehicles,the literature on the integration of uncertain demand and energy consumption of electric vehicles is still ***,practically,it is usually not feasible to ignore the uncertainty of customer demand and the weight-related energy consumption of electronic vehicles(EVs)in actual ***,we propose the robust optimization model based on a route-related uncertain set to tackle this ***,adaptive large neighbourhood search heuristic has been developed to solve the problem due to the NP-hard nature of the *** effectiveness of the method is verified by experiments,and the influence of uncertain demand and uncertain parameters on the solution is further explored.
To gain a deeper understanding of the inherent nature of the difficulty in solving the random regular exact (d，k)-SAT problem，clarify the changing patterns between phase transitions and difficulty，and further desig...
详细信息
Vehicle scheduling plays a profound role in public ***,stochastic vehicle scheduling may lead to more robust *** solve the stochastic vehicle scheduling problem(SVSP),a discrete artificial bee colony algorithm(DABC)is...
详细信息
Vehicle scheduling plays a profound role in public ***,stochastic vehicle scheduling may lead to more robust *** solve the stochastic vehicle scheduling problem(SVSP),a discrete artificial bee colony algorithm(DABC)is *** to the discreteness of SVSP,in DABC,a new encoding and decoding scheme with small dimensions is designed,whilst an initialization rule and three neighborhood search schemes(i.e.,discrete scheme,heuristic scheme,and learnable scheme)are devised individually.A series of experiments demonstrate that the proposed DABC with any neighborhood search scheme is able to produce better schedules than the benchmark results and DABC with the heuristic scheme performs the best among the three proposed search schemes.
A new spotted hyena intelligent optimizer (ISHO) algorithm incorporating multi-strategy improvement was proposed for the characteristics of the capacitated vehicle routing problem (CVRP).A combination of K-means clust...
详细信息
image restoration is a classic foundational visual task, aimed at recovering damaged images, such as those affected by compression, blurring, or noise, to high-definition clarity. Although current image enhancement te...
详细信息
ISBN:
(数字)9798331516147
ISBN:
(纸本)9798331516154
image restoration is a classic foundational visual task, aimed at recovering damaged images, such as those affected by compression, blurring, or noise, to high-definition clarity. Although current image enhancement techniques largely rely on convolutional neural networks, the efficient capabilities of Transformers demonstrated in complex visual tasks have not yet been extensively applied in the field of image restoration. This paper presents an optimized image restoration strategy based on the SwinlR model, which is applied to the task of underwater image restoration. In the shallow feature extraction phase, deep separable convolution technology is employed; while in the deep feature extraction phase, a Spatial Gated Feedforward Network (SGFN) is utilized. The experimental data suggests that the novel approach surpasses conventional techniques in effectiveness and has effectively minimized both the model's parameter quantity and computational demands.
暂无评论