Hasil untuk "Maps"

R. Sánchez-García, J. Gómez-Blanco, A. Cuervo et al.

Cryo-EM maps are valuable sources of information for protein structure modeling. However, due to the loss of contrast at high frequencies, they generally need to be post-processed to improve their interpretability. Most popular approaches, based on global B-factor correction, suffer from limitations. For instance, they ignore the heterogeneity in the map local quality that reconstructions tend to exhibit. Aiming to overcome these problems, we present DeepEMhancer, a deep learning approach designed to perform automatic post-processing of cryo-EM maps. Trained on a dataset of pairs of experimental maps and maps sharpened using their respective atomic models, DeepEMhancer has learned how to post-process experimental maps performing masking-like and sharpening-like operations in a single step. DeepEMhancer was evaluated on a testing set of 20 different experimental maps, showing its ability to reduce noise levels and obtain more detailed versions of the experimental maps. Additionally, we illustrated the benefits of DeepEMhancer on the structure of the SARS-CoV-2 RNA polymerase. Sanchez-Garcia et al. present DeepEMhancer, a deep learning-based method that can automatically perform post-processing of raw cryo-electron microscopy density maps. The authors report that DeepEMhancer globally improves local quality of density maps, and may represent a useful tool for novel structures where PDB models are not readily available.

Deng-Ping Fan, Cheng Gong, Yang Cao et al.

The existing binary foreground map (FM) measures address various types of errors in either pixel-wise or structural ways. These measures consider pixel-level match or image-level information independently, while cognitive vision studies have shown that human vision is highly sensitive to both global information and local details in scenes. In this paper, we take a detailed look at current binary FM evaluation measures and propose a novel and effective E-measure (Enhanced-alignment measure). Our measure combines local pixel values with the image-level mean value in one term, jointly capturing image-level statistics and local pixel matching information. We demonstrate the superiority of our measure over the available measures on 4 popular datasets via 5 meta-measures, including ranking models for applications, demoting generic, random Gaussian noise maps, ground-truth switch, as well as human judgments. We find large improvements in almost all the meta-measures. For instance, in terms of application ranking, we observe improvement ranging from 9.08% to 19.65% compared with other popular measures.

Sasha Babicki, David Arndt, A. Marcu et al.

Heatmapper is a freely available web server that allows users to interactively visualize their data in the form of heat maps through an easy-to-use graphical interface. Unlike existing non-commercial heat map packages, which either lack graphical interfaces or are specialized for only one or two kinds of heat maps, Heatmapper is a versatile tool that allows users to easily create a wide variety of heat maps for many different data types and applications. More specifically, Heatmapper allows users to generate, cluster and visualize: (i) expression-based heat maps from transcriptomic, proteomic and metabolomic experiments; (ii) pairwise distance maps; (iii) correlation maps; (iv) image overlay heat maps; (v) latitude and longitude heat maps and (vi) geopolitical (choropleth) heat maps. Heatmapper offers a number of simple and intuitive customization options for facile adjustments to each heat map's appearance and plotting parameters. Heatmapper also allows users to interactively explore their numeric data values by hovering their cursor over each heat map cell, or by using a searchable/sortable data table view. Heat map data can be easily uploaded to Heatmapper in text, Excel or tab delimited formatted tables and the resulting heat map images can be easily downloaded in common formats including PNG, JPG and PDF. Heatmapper is designed to appeal to a wide range of users, including molecular biologists, structural biologists, microbiologists, epidemiologists, environmental scientists, agriculture/forestry scientists, fish and wildlife biologists, climatologists, geologists, educators and students. Heatmapper is available at http://www.heatmapper.ca.

E. Amir, K. Davis, Michelle D. Tadmor et al.

Ming Liang, Binh Yang, Rui Hu et al.

We propose a motion forecasting model that exploits a novel structured map representation as well as actor-map interactions. Instead of encoding vectorized maps as raster images, we construct a lane graph from raw map data to explicitly preserve the map structure. To capture the complex topology and long range dependencies of the lane graph, we propose LaneGCN which extends graph convolutions with multiple adjacency matrices and along-lane dilation. To capture the complex interactions between actors and maps, we exploit a fusion network consisting of four types of interactions, actor-to-lane, lane-to-lane, lane-to-actor and actor-to-actor. Powered by LaneGCN and actor-map interactions, our model is able to predict accurate and realistic multi-modal trajectories. Our approach significantly outperforms the state-of-the-art on the large scale Argoverse motion forecasting benchmark.

M. Glasser, D. V. Van Essen

Noninvasively mapping the layout of cortical areas in humans is a continuing challenge for neuroscience. We present a new method of mapping cortical areas based on myelin content as revealed by T1-weighted (T1w) and T2-weighted (T2w) MRI. The method is generalizable across different 3T scanners and pulse sequences. We use the ratio of T1w/T2w image intensities to eliminate the MR-related image intensity bias and enhance the contrast to noise ratio for myelin. Data from each subject were mapped to the cortical surface and aligned across individuals using surface-based registration. The spatial gradient of the group average myelin map provides an observer-independent measure of sharp transitions in myelin content across the surface—i.e., putative cortical areal borders. We found excellent agreement between the gradients of the myelin maps and the gradients of published probabilistic cytoarchitectonically defined cortical areas that were registered to the same surface-based atlas. For other cortical regions, we used published anatomical and functional information to make putative identifications of dozens of cortical areas or candidate areas. In general, primary and early unimodal association cortices are heavily myelinated and higher, multimodal, association cortices are more lightly myelinated, but there are notable exceptions in the literature that are confirmed by our results. The overall pattern in the myelin maps also has important correlations with the developmental onset of subcortical white matter myelination, evolutionary cortical areal expansion in humans compared with macaques, postnatal cortical expansion in humans, and maps of neuronal density in non-human primates.

Pingping Zhang, D. Wang, Huchuan Lu et al.

Fully convolutional neural networks (FCNs) have shown outstanding performance in many dense labeling problems. One key pillar of these successes is mining relevant information from features in convolutional layers. However, how to better aggregate multi-level convolutional feature maps for salient object detection is underexplored. In this work, we present Amulet, a generic aggregating multi-level convolutional feature framework for salient object detection. Our framework first integrates multi-level feature maps into multiple resolutions, which simultaneously incorporate coarse semantics and fine details. Then it adaptively learns to combine these feature maps at each resolution and predict saliency maps with the combined features. Finally, the predicted results are efficiently fused to generate the final saliency map. In addition, to achieve accurate boundary inference and semantic enhancement, edge-aware feature maps in low-level layers and the predicted results of low resolution features are recursively embedded into the learning framework. By aggregating multi-level convolutional features in this efficient and flexible manner, the proposed saliency model provides accurate salient object labeling. Comprehensive experiments demonstrate that our method performs favorably against state-of-the-art approaches in terms of near all compared evaluation metrics.

S. Smale

This is a survey article on the area of global analysis defined by differentiable dynamical systems or equivalently the action (differentiable) of a Lie group G on a manifold M. An action is a homomorphism G→Diff(M) such that the induced map G×M→M is differentiable. Here Diff(M) is the group of all diffeomorphisms of M and a diffeo- morphism is a differentiable map with a differentiable inverse. Everything will be discussed here from the C ∞ or C r point of view. All manifolds maps, etc. will be differentiable (C r , 1 ≦ r ≦ ∞) unless stated otherwise.

Jean‐François Rual, K. Venkatesan, Tong Hao et al.

S. Wakana, Hangyi Jiang, Lidia M. Nagae-Poetscher et al.

Dmitry Pokholok, Christopher T. Harbison, S. Levine et al.

P. Menozzi, A. Piazza, L. Cavalli-Sforza

Dominik Kulmer, Maximilian Leitenstern, Marcel Weinmann et al.

Accurate localization is a critical component of mobile autonomous systems, especially in Global Navigation Satellite Systems (GNSS)-denied environments where traditional methods fail. In such scenarios, environmental sensing is essential for reliable operation. However, approaches such as LiDAR odometry and Simultaneous Localization and Mapping (SLAM) suffer from drift over long distances, especially in the absence of loop closures. Map-based localization offers a robust alternative, but the challenge lies in creating and georeferencing maps without GNSS support. To address this issue, we propose a method for creating georeferenced maps without GNSS by using publicly available data, such as building footprints and surface models derived from sparse aerial scans. Our approach integrates these data with onboard LiDAR scans to produce dense, accurate, georeferenced 3D point cloud maps. By combining an Iterative Closest Point (ICP) scan-to-scan and scan-to-map matching strategy, we achieve high local consistency without suffering from long-term drift. Thus, we eliminate the reliance on GNSS for the creation of georeferenced maps. The results demonstrate that LiDAR-only mapping can produce accurate georeferenced point cloud maps when augmented with existing map priors.

Daciberg Lima Goncalves, Robert Skiba, P. Christopher Staecker

This paper concerns various models of ``at-most-$n$-valued maps''. That is, multivalued maps $f:X\multimap Y$ for which $f(x)$ has cardinality at most $n$ for each $x$. We consider 4 classes of such maps which have appeared in the literature: $\mathcal U$, the set of exactly $n$-valued maps, or unions of such; $\mathcal F$, the set of $n$-fold maps defined by Crabb; $\mathcal S$, the set of symmetric product maps; and $\mathcal W$, the set of weighted maps with weights in $\mathbb N$. Our main result is roughly that these classes satisfy the following containments: \[ \mathcal U \subsetneq \mathcal F \subsetneq \mathcal S = \mathcal W \] Furthermore we define the general class $\mathcal C$ of all at-most-$n$-valued maps, and show that there are maps in $\mathcal C$ which are outside of any of the other classes above. We also describe a configuration-space point of view for the class $\mathcal C$, defining a configuration space $C_n(Y)$ such that any at-most-$n$-valued map $f:X\multimap Y$ corresponds naturally to a single-valued map $f:X\to C_n(Y)$. We give a full calculation of the fundamental group and homology groups of $C_n(S^1)$.

Christian Löwens, Thorben Funke, Jingchao Xie et al.

Online mapping models show remarkable results in predicting vectorized maps from multi-view camera images only. However, all existing approaches still rely on ground-truth high-definition maps during training, which are expensive to obtain and often not geographically diverse enough for reliable generalization. In this work, we propose PseudoMapTrainer, a novel approach to online mapping that uses pseudo-labels generated from unlabeled sensor data. We derive those pseudo-labels by reconstructing the road surface from multi-camera imagery using Gaussian splatting and semantics of a pre-trained 2D segmentation network. In addition, we introduce a mask-aware assignment algorithm and loss function to handle partially masked pseudo-labels, allowing for the first time the training of online mapping models without any ground-truth maps. Furthermore, our pseudo-labels can be effectively used to pre-train an online model in a semi-supervised manner to leverage large-scale unlabeled crowdsourced data. The code is available at github.com/boschresearch/PseudoMapTrainer.

Rohan Kumar Mishra, Debidutta Pattnaik, M. Kabir Hassan et al.

This study aims to offer new quantitative and qualitative insights into transaction efficiency and conflict of interest among minority and controlling shareholders in related party transactions (RPTs). We utilize systematic literature review (SLR) and bibliometric techniques to analyse 218 published articles. Our analysis identifies significant contributors, publishing sources, research groups, and maps the evolution of RPT themes and their relationship to contemporary theoretical frameworks. Subsequently, we conduct a comprehensive network and content analysis. Our findings indicate that research in RPTs began evolving post-global financial crisis, particularly since 2008, with East-Asian researchers dominating the intellectual discourse. Most studies are non-collaborative and based on empirical evidence from a limited number of countries. Methodologically, many studies employ descriptive statistics or regression techniques. We identify six thematic clusters contributing to the growth narrative of RPT research. Furthermore, we identify potential avenues for future research in RPTs and corporate governance while highlighting progressive trends and dynamics within the selected themes.

François Siewe, Guy Merlin Ngounou

The unified modelling language (UML) is an industrial de facto standard for system modelling. It consists of a set of graphical notations (also known as diagrams) and has been used widely in many industrial applications. Although the graphical nature of UML is appealing to system developers, the official documentation of UML does not provide formal semantics for UML diagrams. This makes UML unsuitable for formal verification and, therefore, limited when it comes to the development of safety/security-critical systems where faults can cause damage to people, properties, or the environment. The UML activity diagram is an important UML graphical notation, which is effective in modelling the dynamic aspects of a system. This paper proposes a formal semantics for UML activity diagrams based on the calculus of context-aware ambients (CCA). An algorithm (semantic function) is proposed that maps any activity diagram onto a process in CCA, which describes the behaviours of the UML activity diagram. This process can then be executed and formally verified using the CCA simulation tool ccaPL and the CCA runtime verification tool ccaRV. Hence, design flaws can be detected and fixed early during the system development lifecycle. The pragmatics of the proposed approach are demonstrated using a case study in e-commerce.

Yukang Zeng, Zhenhui Shen, Yao Cao et al.

Abstract Due to its asexual reproduction characteristics, Naematelia aurantialba faces limitations in genetic diversity, germplasm identification, and intellectual property protection, necessitating molecular-level approaches to address these challenges. This study utilized resequencing data from 36 germplasm resources of N. aurantialba to conduct a population structure analysis and construct fingerprint maps based on core multinucleotide polymorphisms (MNPs), core single nucleotide polymorphisms (SNPs), and core insertions/deletions (INDELs), while also calculating genetic similarity among strains based on different variant sites. The results indicated that when K = 2, the genetic structure of individual strains was more distinctly divided, a finding corroborated by both the phylogenetic tree and the principal component analysis. A total of 108 core MNPs (comprising 333 SNPs), 54 core SNPs, and 40 core INDELs were identified. Fingerprint maps for the 36 germplasm resources of N. aurantialba were constructed using combinations of different core variant sites. In strains with high similarity, the genetic similarity identified by the three types of core variant sites was all above 97%, and could reach up to 100%, achieving mutual complementary validation. Therefore, it is preliminarily determined that strains with genetic similarity above 97% can be considered as the same strain. These achievements provide valuable resources for the identification of N. aurantialba germplasm and for intellectual property protection. Additionally, the multi-fingerprint map construction method offers new insights for research on other edible fungi.

Saad Khalil, Sol Kim, Bo-In Lee et al.

Depth estimation is crucial for 3D reconstruction and surgical navigation, providing critical insights for endoscopic procedures. While foundation models excel in depth estimation for natural images, their performance in the medical domain remains limited, particularly under challenging conditions like brightness fluctuations. This study develops a robust self-supervised framework for monocular depth estimation to address these challenges. We introduce a layer-wise low-rank adaptation (LW-LoRA) of the Depth-Anything-V2 foundation model, tailored for endoscopic data. Unlike conventional fine-tuning, LW-LoRA assigns an empirically determined rank vector across the encoder layers for efficient training. The method integrates residual convolutional blocks (ResConv) to capture fine-grained details and a multi-head attention-based pose network to enhance camera pose estimation, ensuring accurate 3D reconstructions. A multi-scale SSIM (MS-SSIM) reprojection loss refines depth predictions, while a brightness calibration module ensures robustness against illumination inconsistencies. During training, the backbone encoder is frozen, optimizing only the LoRA layers for efficiency. Extensive evaluations on the SCARED dataset highlight the superior performance of our framework, offering faster inference and high-quality depth maps. Zero-shot testing on Hamlyn and clinical datasets confirms its generalization across diverse data types. Our framework efficiently adapts the foundation model for depth estimation in the medical domain, addressing challenges in endoscopic imaging, such as brightness variations and fine-detail preservation. It enables accurate, dense 3D point cloud reconstructions, ensuring reliable performance in clinical settings.

Peter Jackson