Search

Your search keyword '"Shen Xing"' showing total 2,424 results
Start Over Author "Shen Xing"
2,424 results on '"Shen Xing"'

60. Investigating Factors Affecting Upstream Contamination

Author

Yang, Shen Xing, Hengyue, Wang, Majel, Mendoza Kieran, Ricardo, Bernard, Guo, Huaqun, editor, McLoughlin, Ian, editor, Chekole, Eyasu Getahun, editor, Lakshmanan, Umayal, editor, Meng, Weizhi, editor, Wang, Peng Cheng, editor, and Lu, Jiqiang, editor

Published
2023
Full Text
View/download PDF

70. DCT-Mask: Discrete Cosine Transform Mask Representation for Instance Segmentation

Author

Shen, Xing, Yang, Jirui, Wei, Chunbo, Deng, Bing, Huang, Jianqiang, Hua, Xiansheng, Cheng, Xiaoliang, and Liang, Kewei

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Binary grid mask representation is broadly used in instance segmentation. A representative instantiation is Mask R-CNN which predicts masks on a $28\times 28$ binary grid. Generally, a low-resolution grid is not sufficient to capture the details, while a high-resolution grid dramatically increases the training complexity. In this paper, we propose a new mask representation by applying the discrete cosine transform(DCT) to encode the high-resolution binary grid mask into a compact vector. Our method, termed DCT-Mask, could be easily integrated into most pixel-based instance segmentation methods. Without any bells and whistles, DCT-Mask yields significant gains on different frameworks, backbones, datasets, and training schedules. It does not require any pre-processing or pre-training, and almost no harm to the running speed. Especially, for higher-quality annotations and more complex backbones, our method has a greater improvement. Moreover, we analyze the performance of our method from the perspective of the quality of mask representation. The main reason why DCT-Mask works well is that it obtains a high-quality mask representation with low complexity. Code is available at https://github.com/aliyun/DCT-Mask.git.

Published
2020

71. Deep Euler method: solving ODEs by approximating the local truncation error of the Euler method

Author

Shen, Xing, Cheng, Xiaoliang, and Liang, Kewei

Subjects
Mathematics - Numerical Analysis
Abstract

In this paper, we propose a deep learning-based method, deep Euler method (DEM) to solve ordinary differential equations. DEM significantly improves the accuracy of the Euler method by approximating the local truncation error with deep neural networks which could obtain a high precision solution with a large step size. The deep neural network in DEM is mesh-free during training and shows good generalization in unmeasured regions. DEM could be easily combined with other schemes of numerical methods, such as Runge-Kutta method to obtain better solutions. Furthermore, the error bound and stability of DEM is discussed.

Published
2020

75. A genome-scale phylogeny of the kingdom Fungi

Author

Li, Yuanning, Steenwyk, Jacob L, Chang, Ying, Wang, Yan, James, Timothy Y, Stajich, Jason E, Spatafora, Joseph W, Groenewald, Marizeth, Dunn, Casey W, Hittinger, Chris Todd, Shen, Xing-Xing, and Rokas, Antonis

Subjects
Microbiology, Biological Sciences, Evolutionary Biology, Genetics, Fungi, Genome, Phylogeny, ancient diversification, coalescence, concatenation, phylogenetic signal, phylogenomics, polytomy test, relative evolutionary divergence, taxonomy, zygomycetes, Medical and Health Sciences, Psychology and Cognitive Sciences, Developmental Biology, Biological sciences, Biomedical and clinical sciences, Psychology
Abstract

Phylogenomic studies using genome-scale amounts of data have greatly improved understanding of the tree of life. Despite the diversity, ecological significance, and biomedical and industrial importance of fungi, evolutionary relationships among several major lineages remain poorly resolved, especially those near the base of the fungal phylogeny. To examine poorly resolved relationships and assess progress toward a genome-scale phylogeny of the fungal kingdom, we compiled a phylogenomic data matrix of 290 genes from the genomes of 1,644 species that includes representatives from most major fungal lineages. We also compiled 11 data matrices by subsampling genes or taxa from the full data matrix based on filtering criteria previously shown to improve phylogenomic inference. Analyses of these 12 data matrices using concatenation- and coalescent-based approaches yielded a robust phylogeny of the fungal kingdom, in which ∼85% of internal branches were congruent across data matrices and approaches used. We found support for several historically poorly resolved relationships as well as evidence for polytomies likely stemming from episodes of ancient diversification. By examining the relative evolutionary divergence of taxonomic groups of equivalent rank, we found that fungal taxonomy is broadly aligned with both genome sequence divergence and divergence time but also identified lineages where current taxonomic circumscription does not reflect their levels of evolutionary divergence. Our results provide a robust phylogenomic framework to explore the tempo and mode of fungal evolution and offer directions for future fungal phylogenetic and taxonomic studies.

Published
2021

93. A genome-scale phylogeny of Fungi; insights into early evolution, radiations, and the relationship between taxonomy and phylogeny

Author

Li, Yuanning, Steenwyk, Jacob L, Chang, Ying, Wang, Yan, James, Timothy Y, Stajich, Jason E, Spatafora, Joseph W, Groenewald, Marizeth, Dunn, Casey W, Hittinger, Chris Todd, Shen, Xing-Xing, and Rokas, Antonis

Subjects
Microbiology, Biological Sciences, Evolutionary Biology, Genetics
Abstract

Phylogenomic studies based on genome-scale amounts of data have greatly improved understanding of the tree of life. Despite their diversity, ecological significance, and biomedical and industrial importance, large-scale phylogenomic studies of Fungi are lacking. Furthermore, several evolutionary relationships among major fungal lineages remain controversial, especially those at the base of the fungal phylogeny. To begin filling these gaps and assess progress toward a genome-scale phylogeny of the entire fungal kingdom, we compiled a phylogenomic data matrix of 290 genes from the genomes of 1,644 fungal species that includes representatives from most major fungal lineages; we also compiled 11 additional data matrices by subsampling genes or taxa based on filtering criteria previously shown to improve phylogenomic inference. Analyses of these 12 data matrices using concatenation- and coalescent-based approaches yielded a robust phylogeny of the kingdom in which ∼85% of internal branches were congruent across data matrices and approaches used. We found support for several relationships that have been historically contentious (e.g., for the placement of Wallemiomycotina (Basidiomycota), as sister to Agaricomycotina), as well as evidence for polytomies likely stemming from episodes of ancient diversification (e.g., at the base of Basidiomycota). By examining the relative evolutionary divergence of taxonomic groups of equivalent rank, we found that fungal taxonomy is broadly aligned with genome sequence divergence, but also identified lineages, such as the subphylum Saccharomycotina, where current taxonomic circumscription does not fully account for their high levels of evolutionary divergence. Our results provide a robust phylogenomic framework to explore the tempo and mode of fungal evolution and directions for future fungal phylogenetic and taxonomic studies.

Published
2020

Catalog

Books, media, physical & digital resources
See catalog results