Though widely used in image classification, convolutional neural networks (CNNs) are prone to noise interruptions, i.e. the CNN output can be drastically changed by small image noise. To improve the noise robustness, we try to integrate CNNs with wavelet by replacing the common down-sampling (maxpooling, strided-convolution, and average pooling) with discrete wavelet transform (DWT). We firstly propose general DWT and inverse DWT (IDWT) layers applicable to various orthogonal and biorthogonal discrete wavelets like Haar, Daubechies, and Cohen, etc., and then design wavelet integrated CNNs (WaveCNets) by integrating DWT into the commonly used CNNs (VGG, ResNets, and DenseNet). During the down-sampling, WaveCNets apply DWT to decompose the feature maps into the low-frequency and high-frequency components. Containing the main information including the basic object structures, the low-frequency component is transmitted into the following layers to generate robust high-level features. The high-frequency components are dropped to remove most of the data noises. The experimental results show that WaveCNets achieve higher accuracy on ImageNet than various vanilla CNNs. We have also tested the performance of WaveCNets on the noisy version of ImageNet, ImageNet-C and six adversarial attacks, the results suggest that the proposed DWT/IDWT layers could provide better noise-robustness and adversarial robustness. When applying WaveCNets as backbones, the performance of object detectors (i.e., faster R-CNN and RetinaNet) on COCO detection dataset are consistently improved. We believe that suppression of aliasing effect, i.e. separation of low frequency and high frequency information, is the main advantages of our approach. The code of our DWT/IDWT layer and different WaveCNets are available at https://github.com/CVI-SZU/WaveCNet.
In deep networks, the lost data details significantly degrade the performances of image segmentation. In this paper, we propose to apply Discrete Wavelet Transform (DWT) to extract the data details during feature map down-sampling, and adopt Inverse DWT (IDWT) with the extracted details during the up-sampling to recover the details. We firstly transform DWT/IDWT as general network layers, which are applicable to 1D/2D/3D data and various wavelets like Haar, Cohen, and Daubechies, etc. Then, we design wavelet integrated deep networks for image segmentation (WaveSNets) based on various architectures, including U-Net, SegNet, and DeepLabv3+. Due to the effectiveness of the DWT/IDWT in processing data details, experimental results on CamVid, Pascal VOC, and Cityscapes show that our WaveSNets achieve better segmentation performances than their vanilla versions.
Background Variations in human papillomavirus (HPV) E6 and E7 have been shown to be closely related to the persistence of the virus and the occurrence and development of cervical cancer. Long control region (LCR) of HPV has been shown multiple functions on regulating viral transcription. In recent years, there have been reports on E6/E7/LCR of HPV-16 and HPV-58, but there are few studies on HPV-52, especially for LCR. In this study, we focused on gene polymorphism of the HPV-52 E6/E7/LCR sequences, assessed the effects of variations on the immune recognition of viral E6 and E7 antigens, predicted the effect of LCR variations on transcription factor binding sites and provided more basic date for further study of E6/E7/LCR in Chengdu, China. Methods LCR/E6/E7 of the HPV-52 were amplified and sequenced to do polymorphic and phylogenetic analysis. Sequences were aligned with the reference sequence by MEGA 7.0 to identify SNP. A neighbor-joining phylogenetic tree was constructed by MEGA 7.0, followed by the secondary structure prediction of the related proteins using PSIPRED 4.0. The selection pressure of E6 and E7 coding regions were estimated by Bayes empirical Bayes analysis of PAML 4.9. The HLA class-I and II binding peptides were predicted by the Immune Epitope Database server. The B cell epitopes were predicted by ABCpred server. Transcription factor binding sites in LCR were predicted by JASPAR database. Results 50 SNP sites (6 in E6, 10 in E7, 34 in LCR) were found. From the most variable to the least variable, the nucleotide variations were LCR > E7 > E6. Two deletions were found between the nucleotide sites 7387–7391 (TTATG) and 7698–7700 (CTT) in all samples. A deletion was found between the nucleotide sites 7287–7288 (TG) in 97.56% (40/41) of the samples. The combinations of all the SNP sites and deletions resulted in 12 unique sequences. As shown in the neighbor-joining phylogenetic tree, except for one belonging to sub-lineage C2, others sequences clustered into sub-lineage B2. No positive selection was observed in E6 and E7. 8 non-synonymous amino acid substitutions (including E3Q and K93R in the E6, and T37I, S52D, Y59D, H61Y, D64N and L99R in the E7) were potential affecting multiple putative epitopes for both CD4+ and CD8+ T-cells and B-cells. A7168G was the most variable site (100%) and the binding sites for transcription factor VAX1 in LCR. In addition, the prediction results showed that LCR had the high probability binding sites for transcription factors SOX9, FOS, RAX, HOXA5, VAX1 and SRY. Conclusion This study provides basic data for understanding the relation among E6/E7/LCR mutations, lineages and carcinogenesis. Furthermore, it provides an insight into the intrinsic geographical relatedness and biological differences of the HPV-52 variants, and contributes to further research on the HPV-52 therapeutic vaccine development.
This paper constructs a set partition coding system (SPACS) to combine the advantages of different types of set partition coding algorithms. General tree (GT) is an important conception introduced in this paper, which can represent tree set and square set simultaneously. With the help of GT, SPIHT is generalized to construct degree- k SPIHT based on the analysis of two kinds of set partition operations. Using the same coding mechanism, SPACS (k,p) is constructed, aided with virtual subbands that are generated by recursive division on the LL band. SPACS belongs to tree-set partition coding algorithms if k and p take smaller values. In particular, SPACS(2,1) is the classical SPIHT. SPACS tends toward a block-set partition coding algorithm as k,p increases. Location bit, amplitude bit, and unnecessary bit are presented, which can be used to analyze the coding efficiency of SPACS. We compress 256 images with 512×512 using SPACS. The numerical results show SPACS achieves some improvements in coding efficiency over SPIHT, especially at very low bitrate. On average, to code every image, SPACS(3,1) (at an average of 3.93 bpp) needs 7792 more location bits but saves 10 218 unnecessary bits, compared with SPIHT (3.94 bpp).
Artocarpus nanchuanensis (Moraceae), which is naturally distributed in China, is a representative and extremely endangered tree species. In this study, we obtained a high-quality chromosome-scale genome assembly and annotation information for A. nanchuanensis using integrated approaches, including Illumina, Nanopore sequencing platform, and Hi-C. A total of 128.71 Gb of raw Nanopore reads were generated from 20-kb libraries, and 123.38 Gb of clean reads were obtained after filtration with 160.34× coverage depth and a 17.48-kb average read length. The final assembled A. nanchuanensis genome was 769.44 Mb with a 2.09 Mb contig N50, and 99.62% (766.50 Mb) of the assembled data was assigned to 28 pseudochromosomes. In total, 39,596 genes (95.10%, 39,596/41,636) were successfully annotated, and 129 metabolic pathways were detected. Plants disease resistance/insect resistance genes, plant–pathogen interaction metabolic pathways, and abundant biosynthesis pathways of vitamins, flavonoid, and gingerol were detected. Unigene reveals the basis of species-specific functions, and gene family in contraction and expansion generally implies strong functional differences in the evolution. Compared with other related species, a total of 512 unigenes, 309 gene families in contraction, and 559 gene families in expansion were detected in A. nanchuanensis. This A. nanchuanensis genome information provides an important resource to expand our understanding of the unique biological processes, nutritional and medicinal benefits, and evolutionary relationship of this species. The study of gene function and metabolic pathway in A. nanchuanensis may reveal the theoretical basis of a special trait in A. nanchuanensis and promote the study and utilization of its rare medicinal value.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.