The integration of psychology and computer science has become the mainstream contemporary research method on psychological data. Weibo, China's largest open platform for communication and information sharing betwe...
详细信息
Underwater optical imaging produces images with high resolution and abundant information and hence has outstanding advantages in short-distance underwater target ***,low-light and high-noise scenarios pose great chall...
详细信息
Underwater optical imaging produces images with high resolution and abundant information and hence has outstanding advantages in short-distance underwater target ***,low-light and high-noise scenarios pose great challenges in un-derwater image and video *** improve the accuracy and anti-noise performance of underwater target image edge detection,an underwater target edge detection method based on ant colony optimization and reinforcement learning is proposed in this ***,the reinforcement learning concept is integrated into artificial ants’movements,and a variable radius sensing strategy is pro-posed to calculate the transition probability of each *** methods aim to avoid undetection and misdetection of some pixels in image ***,a double-population ant colony strategy is proposed,where the search process takes into account global search and local search *** results show that the algorithm can effectively extract the contour information of underwater targets and keep the image texture well and also has ideal anti-interference performance.
In this paper, we take Mongolian ethnic patterns as the research object and conduct a technical study on ethnic pattern generation and super-resolution reconstruction based on deep learning generative adversarial netw...
In this paper, we take Mongolian ethnic patterns as the research object and conduct a technical study on ethnic pattern generation and super-resolution reconstruction based on deep learning generative adversarial networks to address the problems of low quality of ethnic pattern data collection and lack of innovation. In this paper, we use a database including 1621 Mongolian ethnic patterns, train StyleGAN2 on the pre-processed images, and generate images After that, we use ESRGAN to perform super-resolution reconstruction on the generated images to generate high-resolution patterns with Mongolian style. The model designed in this paper can generate high quality images with ethnic characteristics in the case of insufficient original data. Compared with the time-consuming and labor-intensive traditional ethnic pattern design methods, the model designed in this paper lowers the threshold of ethnic pattern innovation and contributes to the innovative design of ethnic patterns to a certain extent, which has some positive significance for the protection and inheritance of ethnic patterns.
Every univariate Hermite interpolation problem can be written as a pointwise limit of Lagrange ***,this property is not preserved for the multivariate *** this paper,the authors first generalize the result of *** an a...
详细信息
Every univariate Hermite interpolation problem can be written as a pointwise limit of Lagrange ***,this property is not preserved for the multivariate *** this paper,the authors first generalize the result of *** an application,the authors consider the discrete approximation problem for a special case when the interpolation condition contains all partial derivatives of order less than n and one nth order differential *** addition,for the case of n≥3,the authors use the concept of Cartesian tensors to give a sufficient condition to find a sequence of discrete points,such that the Lagrange interpolation problems at these points converge to the given Hermite-type interpolant.
Despite the great success of spoken language understanding (SLU) in high-resource languages, it remains challenging in low-resource languages mainly due to the lack of labeled training data. The recent multilingual co...
详细信息
Supervised methods on 3D medical image segmentation need large amounts of annotated data, but annotating is time-consuming. Also, existing 3D segmentation methods capture more global structural information but overloo...
详细信息
Recently, there has been a surge in recommendations based on heterogeneous information networks (HINs), attributed to their ability to integrate complex and rich semantics. Despite this advancement, most HIN-based rec...
详细信息
With the implementation of the "Internet+" strategy, electronic medical records are generally applied in the medical field. Deep mining of electronic medical record content data is an effective means to obta...
With the implementation of the "Internet+" strategy, electronic medical records are generally applied in the medical field. Deep mining of electronic medical record content data is an effective means to obtain medical knowledge and analyse patients' states, but the existing methods for extracting entities from electronic medical records have problems of redundant information, overlapping entities, and low accuracy rates. Therefore, this paper proposes an entity extraction method for electronic medical records based on the network framework of BERT-Bi LSTM, which incorporates a multichannel self-attention mechanism and location relationship features. First, the text input sequence was encoded using the BERT-BiLSTM network framework, and the global semantic information of the sentence was mined more deeply using the multichannel self-attention mechanism. Then, the position relation characteristic was used to extract the local semantic message of the text, and the position relation characteristic of the word and the position embedding matrix of the whole sentence were obtained. Next, the extracted global semantic information was stitched with the positional embedding matrix of the sentence to obtain the current entity classification matrix. Finally,the proposed method was validated on the dataset of Chinese medical text entity relationship extraction and the 2010i2b2/VA relationship corpus, and the experimental results indicate that the proposed method surpasses existing methods in terms of precision, recall, F1 value and training time.
Text classification is a crucial and fundamental task in natural language processing. Compared with the previous learning paradigm of pre-training and fine-tuning by cross entropy loss, the recently proposed supervise...
详细信息
Regarding computer security, the growth of code vulnerability types presents a persistent challenge. These vulnerabilities, which may cause severe consequences, necessitate precise classification for effective mitigat...
详细信息
暂无评论