Experimental environment and parameter setting
Experimental environment
The experiments were conducted using Python 3.8 with PyTorch as the neural network framework32, implemented in the PyCharm IDE. The details of the computer hardware platform…
Category: 7. Maths
-
Lightweight underwater object detection method based on multi-scale edge information selection
Data augmentation
Under underwater conditions, the limitations in both the quantity and quality of data often impact the model’s performance and generalization ability. Therefore, data augmentation methods are necessary to mitigate the effects…
Continue Reading
-
A multimodal deep learning architecture for predicting interstitial glucose for effective type 2 diabetes management
Study protocol
This study is part of the GATEKEEPER strategy for the Multinational Large-Scale Piloting of an eHealth Platform23. The data were collected in the frame of the Central Greece High Complexity Phase I pilot study: a non-interventional,…
Continue Reading
-
Nonlinear behavior of dispersive solitary wave solutions for the propagation of shock waves in the nonlinear coupled system of equations
The wave transformation for Eq.(1), consider as
$$\begin{aligned} U ( x, t )= W( \varsigma ), \quad V ( x, t )= Q( \varsigma ), \quad \quad \varsigma = x + \mu t. \end{aligned}$$
(17)
Substitute Eq.(17) into…
Continue Reading
-
Method for extracting ship shaft rate features by fusing acoustic and magnetic field
Simultaneous acquisition of shaft frequency information from both acoustic and magnetic fields enhances the identification and assessment of the target. This is because acoustic and magnetic signals reflect the characteristics of the ship target…
Continue Reading
-
Sampling zero stability in sampled data control systems with delays using backward triangle sample and hold
In this part of the section, we introduce a novel polynomial that emerges during the discretization process of time-delay systems utilizing the BTSH method. This polynomial, denoted as \({B_{B,r}}(z,f,\Delta )\), plays a central role in…
Continue Reading
-
Sign language recognition based on dual-channel star-attention convolutional neural network
This experiment is primarily divided into two phases: the first phase involves dataset preprocessing, and the second phase encompasses model construction and training. After simple data preprocessing, the data is fed into the SACNN model for…
Continue Reading
-
Pretraining-improved Spatiotemporal graph network for the generalization performance enhancement of traffic forecasting
Although many advanced traffic prediction models, such as those based on GCN and Spatiotemporal Graph Neural Networks (STGNN), have demonstrated promising results in specific scenarios, they often struggle to adapt to diverse traffic networks and…
Continue Reading
-
Visual delta generation with large multi-modal models enhances composed image retrieval using unlabeled data
Sec. 5.1 details the setup and VDG generation and refer this section for data availability. Sec. 5.2 covers quality checks of visual deltas. Evaluations and baseline comparisons are in Sec. 5.3, and further analyses in Sec. 5.6.
Continue Reading
-
Location allocation approach based on influence of points of dispensing
This section presents existing work on locating new facilities/PODs where facilities/PODs already exist based on location selection queries and facility location problems.
Location selection query
The location selection query is filed in a spatial…
Continue Reading