Automated Scan-vs-BIM Registration Using Columns Segmented by Deep Learning for Construction Progress Monitoring [PDF]
In construction automation applications, coarse registration between 3D Building Information Modelling (BIM) and the as-built point cloud is vital for the monitoring of construction progress.
G. Z. Tsige +3 more
doaj +1 more source
Adversarial Deformation Regularization for Training Image Registration Neural Networks [PDF]
We describe an adversarial learning approach to constrain convolutional neural network training for image registration, replacing heuristic smoothness measures of displacement fields often used in these tasks.
Barratt, Dean C. +8 more
core +2 more sources
Coupled non-parametric shape and moment-based inter-shape pose priors for multiple basal ganglia structure segmentation [PDF]
This paper presents a new active contour-based, statistical method for simultaneous volumetric segmentation of multiple subcortical structures in the brain.
Cetin, Mujdat +11 more
core +3 more sources
ENFT: Efficient Non-Consecutive Feature Tracking for Robust Structure-from-Motion
Structure-from-motion (SfM) largely relies on feature tracking. In image sequences, if disjointed tracks caused by objects moving in and out of the field of view, occasional occlusion, or image noise, are not handled well, corresponding SfM could be ...
Bao, Hujun +5 more
core +1 more source
Robust Non-Rigid Registration with Reweighted Position and Transformation Sparsity [PDF]
Non-rigid registration is challenging because it is ill-posed with high degrees of freedom and is thus sensitive to noise and outliers. We propose a robust non-rigid registration method using reweighted sparsities on position and transformation to ...
Guo, Daoliang +3 more
core +3 more sources
In situ molecular organization and heterogeneity of the Legionella Dot/Icm T4SS
We present a nearly complete in situ model of the Legionella Dot/Icm type IV secretion system, revealing its central secretion channel and identifying new components. Using cryo‐electron tomography with AI‐based modeling, our work highlights the structure, variability, and mechanism of this complex nanomachine, advancing understanding of bacterial ...
Przemysław Dutka +11 more
wiley +1 more source
Optimization Beyond the Convolution: Generalizing Spatial Relations with End-to-End Metric Learning
To operate intelligently in domestic environments, robots require the ability to understand arbitrary spatial relations between objects and to generalize them to objects of varying sizes and shapes. In this work, we present a novel end-to-end approach to
Abdo, Nichola +3 more
core +1 more source
Cell wall target fragment discovery using a low‐cost, minimal fragment library
LoCoFrag100 is a fragment library made up of 100 different compounds. Similarity between the fragments is minimized and 10 different fragments are mixed into a single cocktail, which is soaked to protein crystals. These crystals are analysed by X‐ray crystallography, revealing the binding modes of the bound fragment ligands.
Kaizhou Yan +5 more
wiley +1 more source
Dual-quaternion-based iterative algorithm of the three dimensional coordinate transformation
Nowadays a unit quaternion is widely employed to represent the three-dimensional (3D) rotation matrix and then applied to the 3D similarity coordinate transformation.
Huaien Zeng +5 more
doaj +1 more source
Applying transformation parameters for 3D transformation between systems WGS 84 and RES (B) [PDF]
Coordinate transformation between 3D cartesian systems: the Reference Elipsoidical System for Bessels Ellipsoid - RES(B) (with coordinates C R(B)), and the World geodetic system WGS 84 (with coordinates C w) are used for transformations C W→ C R(
adera Milan +4 more
doaj

