Recently, there has been an increasing interest in deep graph similarity learning, where the key idea is to learn a deep learning model that maps input graphs to a target space such that the distance in the target space approximates the structural distance in the input space. Here, we ...
Zhang Z, Cui P, Zhu W. Deep learning on graphs: A survey[J]. IEEE Transactions on Knowledge and Data Engineering, 2020. 18年的一篇GNN综述,读完之后,感觉GCN那一部分对我帮助还不小,帮我理清了脉络,也可能是因为之前把《Graph Representation Learning》这本书看完了,所以阅读过程还比较顺利。后面的VG...
Deep learning technology has enabled successful modeling of complex facial features when high-quality images are available. Nonetheless, accurate modeling and recognition of human faces in real-world scenarios “on the wild” or under adverse conditions remains an open problem. Consequently, a plethora ...
which aims to project unimodal representations together into a shared semantic subspace such that the multimodal features can be fused; (ii) coordinated representation including cross-modal similarity models and canonical correlation analysis, which seeks to learn separated but constrained...
Self-supervised Learning(自监督学习):无监督学习的子集,用自动生成的标签对ConvNets进行明确训练。本文主要研究了基于卷积网络的视觉特征学习的自监督学习方法,该方法可以将特征迁移到多个不同的计算机视觉任务中。 由于在自我监督训练期间不需要人工标注来生成伪标签,因此可以使用非常大规模的数据集进行自我监督训练。在...
Step 1: The similarity between the query and each key is calculated to obtain the weight. Common similarity functions include the dot product, concatenating, and perceptron. The related descriptions are described as follows:(26)f(Q,K)=QTK,dotQTWaK,generalWa[Q;K],concatvattanh(WaQ+UaK),per...
The field of graph deep learning is still rapidly evolving and many research ideas emerge by standing on the shoulders of giants. To ease the process,DGl-Gois a command-line interface to get started with training, using and studying state-of-the-art GNNs. DGL collects a rich set ofexample...
Deep Learning for 3D Point Clouds: A Survey 论文阅读 Abstract:在点云深度学习中,主要包含的任务有:3D形状分类、3D目标检测和跟踪、3D点云分割。 Introduction:3D数据通常有许多种表现形式:深度图、点云、网格、体积网格(volumetric grids)。点云表示的好处是:保持了最原始的3D空间中的几何信息,并且没有任何的...
In particular, we design the search string so that only publications containing the three main concepts relevant for this survey are retrieved: log data, anomaly detection, and deep learning. Since some publications use different terminology and to decrease the likelihood that relevant publications are...
[197]作为先驱性的工作,首次提出了Similarity Group Proposal Network(SGPN)。该方法首先对每个点学习特征和语义map,接着引入相似度矩阵来表示各对点之间的相似度。为了学习到更多的判别式特征,使用了double-hinge loss来互相适应相似度矩阵和语义分割的结果。最后使用启发式的NMS方法将相似的点归并进一个实例中。由于...