Shape Classification Research Articles

Stability of rotating equilibrium states of fluid deformable surfaces

AbstractWe consider rotating equilibrium states of fluid deformable surfaces. These states are characterized by a force balance between centrifugal and bending forces, involve surface Killing vector fields and are independent of surface viscosity. Considering a continuum description based on the incompressible surface Navier‐Stokes equations with bending forces and conserved enclosed volume we numerically demonstrate that these rotating equilibrium states can be reached, but also that these states are not stable. Any perturbation in shape or rotating flow field leads to dissipation and destroys the rotating equilibrium states. After breaking symmetry, evolution reaches other rotating states with a lower energy for which the symmetry axis and the rotation axis are not aligned. Such flow fields could be characterized by three‐dimensional Killing vector fields. However, also these states are also not stable. Based on these numerical results we postulate a cascading mechanism of ‘disturbance ‐ force balance reconfiguration ‐ dissipation’ that contains various rotating equilibrium states as transient configurations but eventually leads to the classical equilibrium shapes of the Helfrich energy.

A viewpoint-guided prototype network for 3D shape classification

A viewpoint-guided prototype network for 3D shape classification

Distinguishing and Matching-Aware Unsupervised Point Cloud Completion

Real-scanned point clouds are often incomplete due to occlusion, light reflection and limitations of sensor resolution, which impedes the related progress of downstream tasks, e.g., shape classification and object detection. Although there has been impressive research progress on the point cloud completion topic, they rely on the premise of extensive paired training data. However, collecting complete point clouds in some specified scenarios is labor-intensive and even impractical. To mitigate this problem, we propose DMNet, a distinguishing and matching-aware unsupervised point cloud completion network. Our work belongs to the group of unsupervised completion methods but goes beyond previous studies. Firstly, we propose a distinguishing-aware feature extractor to learn discriminable semantic information for different instances, simultaneously enhancing the robust invariant representation under noise disturbances. Secondly, we design a hierarchy-aware hyperbolic decoder to recover the complete geometry of point clouds, which not only can capture the implicit hierarchical relationships in data but also has an explicit extended nature. Finally, we develop a matching-aware refiner to eliminate noise points via aligning the topology structure of the input and predicted partial point clouds. Extensive experiments on MVP, Completion3D and KITTI datasets prove the effectiveness of our method, which performs favorably over state-of-the-art methods both quantitatively and qualitatively.

Coarse Aggregate Shape Classification Method Based on Per-Optuna-LightGBM Model

To improve the detection level of aggregate shape for automated road use, Per-Optuna-LightGBM model for aggregate shape classification is proposed. Collect aggregate images using industrial camera and extract 48 morphological feature parameters. A feature importance analysis method based on Spearman Correlation and Permutation Importance is proposed to remove redundant factors and select the feature parameters of aggregate morphology. Based on cross-validation, an optimized Optuna-LightGBM model is trained based on the constructed dataset. Compared with GS-XGBoost algorithm, the Optuna-LightGBM model can classify the shape of aggregates more accurately and efficiently. The accuracy value of the proposed model is 82.5%, which increased by 4% compared to before optimization. The proposed model can efficiently classify the shape of aggregates which meet the design requirements, also provide a certain foundation for automated classification of aggregate shapes.

LCPFormer: Towards Effective 3D Point Cloud Analysis via Local Context Propagation in Transformers

Transformer with its underlying attention mechanism and the ability to capture long-range dependencies makes it become a natural choice for unordered point cloud data. However, local regions separated from the general sampling architecture corrupt the structural information of the instances, and the inherent relationships between adjacent local regions lack exploration. In other words, the transformer only focuses on the long-range dependence, while local structural information is still crucial in a transformer-based 3D point cloud model. To enable transformers to incorporate local structural information, we proposed a straightforward solution based on the natural structure of the point clouds to exploit the message passing between neighboring local regions, thus making their representations more comprehensive and discriminative. Concretely, the proposed module, named Local Context Propagation (LCP), is inserted between two transformer layers. It takes advantage of the overlapping points of adjacent local regions (statistically shown to be prevalent) as intermediaries, then re-weighs the features of these shared points from different local regions before passing them to the next layers. Finally, we design a flexible LCPFormer architecture equipped with the LCP module, which is applicable to several different tasks. Experimental results demonstrate that our proposed LCPFormer outperforms various transformer-based methods in benchmarks including 3D shape classification and dense prediction tasks such as 3D object detection and semantic segmentation. Code will be released for reproduction.

3D reconstruction based on hierarchical reinforcement learning with transferability

3D reconstruction is extremely important in CAD (computer-aided design)/CAE (computer-aided Engineering)/CAM (computer-aided manufacturing). For interpretability, reinforcement learning (RL) is used to reconstruct 3D shapes from images by a series of editing actions. However, typical applications of RL for 3D reconstruction face problems. The search space will increase exponentially with the action space due to the curse of dimensionality, which leads to low performance, especially for complex action spaces in 3D reconstruction. Additionally, most works involve training a specific agent for each shape class without learning related experiences from others. Therefore, we present a hierarchical RL approach with transferability to reconstruct 3D shapes (HRLT3D). First, actions are grouped into macro actions that can be chosen by the top-agent. Second, the task is accordingly decomposed into hierarchically simplified sub-tasks solved by sub-agents. Different from classical hierarchical RL (HRL), we propose a sub-agent based on augmented state space (ASS-Sub-Agent) to replace a set of sub-agents, which can speed up the training process due to shared learning and having fewer parameters. Furthermore, the ASS-Sub-Agent is more easily transferred to data of other classes due to the augmented diverse states and the simplified tasks. The experimental results on typical public dataset show that the proposed HRLT3D performs overwhelmingly better than recent baselines. More impressingly, the experiments also demonstrate the extreme transferability of our approach among data of different classes.

An improved fused feature residual network for 3D point cloud data.

Point clouds have evolved into one of the most important data formats for 3D representation. It is becoming more popular as a result of the increasing affordability of acquisition equipment and growing usage in a variety of fields. Volumetric grid-based approaches are among the most successful models for processing point clouds because they fully preserve data granularity while additionally making use of point dependency. However, using lower order local estimate functions to close 3D objects, such as the piece-wise constant function, necessitated the use of a high-resolution grid in order to capture detailed features that demanded vast computational resources. This study proposes an improved fused feature network as well as a comprehensive framework for solving shape classification and segmentation tasks using a two-branch technique and feature learning. We begin by designing a feature encoding network with two distinct building blocks: layer skips within, batch normalization (BN), and rectified linear units (ReLU) in between. The purpose of using layer skips is to have fewer layers to propagate across, which will speed up the learning process and lower the effect of gradients vanishing. Furthermore, we develop a robust grid feature extraction module that consists of multiple convolution blocks accompanied by max-pooling to represent a hierarchical representation and extract features from an input grid. We overcome the grid size constraints by sampling a constant number of points in each grid using a simple K-points nearest neighbor (KNN) search, which aids in learning approximation functions in higher order. The proposed method outperforms or is comparable to state-of-the-art approaches in point cloud segmentation and classification tasks. In addition, a study of ablation is presented to show the effectiveness of the proposed method.

Deep learning model with collage images for the segmentation of dedicated breast positron emission tomography images.

Dedicated breast positron emission tomography (dbPET) has high contrast and resolution optimized for detecting small breast cancers, leading to its noisy characteristics. This study evaluated the application of deep learning to the automatic segmentation of abnormal uptakes on dbPET to facilitate the assessment of lesions. To address data scarcity in model training, we used collage images composed of cropped abnormal uptakes and normal breasts for data augmentation. This retrospective study included 1598 examinations between April 2015 and August 2020. A U-Net-based model with an uptake shape classification head was trained using either the original or augmented dataset comprising collage images. The Dice score, which measures the pixel-wise agreement between a prediction and its ground truth, of the models was compared using the Wilcoxon signed-rank test. Moreover, the classification accuracies were evaluated. After applying the exclusion criteria, 662 breasts were included; among these, 217 breasts had abnormal uptakes (mean age: 58 ± 14years). Abnormal uptakes on the cranio-caudal and mediolateral maximum intensity projection images of 217 breasts were annotated and labeled as focus, mass, or non-mass. The inclusion of collage images into the original dataset yielded a Dice score of 0.884 and classification accuracy of 91.5%. Improvement in the Dice score was observed across all subgroups, and the score of images without breast cancer improved significantly from 0.750 to 0.834 (effect size: 0.76, P = 0.02). Deep learning can be applied for the automatic segmentation of dbPET, and collage images can improve model performance.

Comprehensive Study on Shape Representation Methods for Shape-Based Object Recognition

This study presents a thorough analysis of shape representation techniques. Furthermore, a taxonomy of strategies for form representation and description is provided. This study aims to review the current achievements comprehensively, highlight the weaknesses and advantages of various existing methods in shape representation methods, addressing current research issues and challenging tasks in this research scope. The contributions of this study lie in providing a comprehensive analysis of shape representation techniques, along with a taxonomy of strategies for form representation and description. The research explores the application of shape border information and inner content in various domains, such as form retrieval, shape classification, and medical image analysis. The novelties include the systematic categorization of methods, addressing current research issues, and offering insights into the benefits and drawbacks of existing shape representation approaches, thus paving the way for potential advancements in the field. In the end, the performance of existing methods is discussed to address the drawbacks and effectiveness of current methods.

Revisiting the Cephalic Index: The Origin, Purpose, and Current Applicability—A Narrative Review

ABSTRACT Introduction Cephalometry is the measurement and study of the proportions of the head and face, and infant skull deformities have been the subject of many clinical studies over the last 30 years. Clinical protocols to document change are especially critical during periods of growth and development, but reference values and norms are lacking for infants, especially in the first year after birth. Discussion Key publications were reviewed by the authors to provide a narrative review of the 1) origin of the cephalic index (CI), 2) adaptations of the measure over time, 3) anatomical landmarks and alignments used, 4) clinical tools, 5) necessary considerations for different ethnic, sex, and cultural influences, 6) value and limitations of CI measurements, and 7) appropriate use of the CI in the documentation of infant skull deformities. The CI calculation has remained constant since the 1840s, although the cranial shape classifications have changed many times over the years. Conclusions Used in isolation, the two-dimensional (2D) and linear nature of the CI fails to describe the magnitude of the three-dimensional (3D) head deformity. However, the CI can and should be used in conjunction with other 2D and 3D measurements to provide a more detailed description and understanding of the entire cranial deformity. For today's cranial clinicians and allied health care professionals, the commonly referenced CI chart fails to reflect 1) the natural skull changes noted with supine sleeping infants, and 2) the cultural diversity of the infant population currently treated. Although the CI provides information about one of the most obvious features of the dysmorphology, clinicians must also document and report other distinct features such as frontal flattening or bossing, parietal shape or involvement, sloping, displacement, and/or posterior asymmetry. Clinical Relevance The CI should be used in conjunction with other 2D and 3D anthropometric measurements to provide a more detailed assessment of the entire cranial deformity. The CI results should be discussed with the medical team along with a complete review of other 2D and 3D measurements to determine the best course of action for orthotic treatment or nontreatment of the head shape.

Shapes of Hyperspectral Imaged Microplastics.

Shape matters for microplastics, but its definition, particularly for hyperspectral imaged microplastics, remains ambiguous and inexplicit, leading to incomparability across data. Hyperspectral imaging is a common approach for quantification, yet no unambiguous microplastic shape classification exists. We conducted an expert-based survey and proposed a set of clear and concise shapes (fiber, rod, ellipse, oval, sphere, quadrilateral, triangle, free-form, and unidentifiable). The categories were validated on images of 11,042 microplastics from four environmental compartments (seven matrices: indoor air; wastewater influent, effluent, and sludge; marine water; stormwater; and stormwater pond sediments), by inviting five experts to score each shape. We found that the proposed shapes were well defined, representative, and distinguishable to the human eye, especially for fiber and sphere. Ellipse, oval, and rod were though less distinguishable but dominated in all water and solid matrices. Indoor air held more unidentifiable, an abstract shape that appeared mostly for particles below 30 μm. This study highlights the need for assessing the recognizability of chosen shape categories prior to reporting data. Shapes with a clear and stringent definition would increase comparability and reproducibility across data and promote harmonization in microplastic research.

Variational Pruning of Medial Axes of Planar Shapes

AbstractMedial axis (MA) is a classical shape descriptor in graphics and vision. The practical utility of MA, however, is hampered by its sensitivity to boundary noise. To prune unwanted branches from MA, many definitions of significance measures over MA have been proposed. However, pruning MA using these measures often comes at the cost of shrinking desirable MA branches and losing shape features at fine scales. We propose a novel significance measure that addresses these shortcomings. Our measure is derived from a variational pruning process, where the goal is to find a connected subset of MA that includes as many points that are as parallel to the shape boundary as possible. We formulate our measure both in the continuous and discrete settings, and present an efficient algorithm on a discrete MA. We demonstrate on many examples that our measure is not only resistant to boundary noise but also excels over existing measures in preventing MA shrinking and recovering features across scales.

Anti-Alzheimer Drug Delivery via Pickering Emulsions Stabilized by Plate-like Cellulose Nanocrystals.

In this work, we studied for the first time the formation of olive oil emulsions in water stabilized by plate-like nanocrystals with the supramolecular structure of cellulose II (pCNC). Effects of storage, pCNC concentration, and NaCl on the stability and properties of Pickering emulsions, including the creaming index, droplet size, zeta potential, acid-base surface properties, and rheology, were studied. A significant influence of the shape of nanoparticles (compared to the classical rod-like shape) on the stability parameters and rheological characteristics of emulsions is shown. Plate-like cellulose nanocrystals at a concentration of 16 g/L are able to form delamination-resistant emulsions without added electrolytes. The viscosity of pCNC-stabilized emulsions tends to decrease with increasing electrolyte concentration in the system, which is not characteristic of rod-like CNC-stabilized emulsions. This effect in pCNC-stabilized emulsions assumedly can be associated both with weak mechanical engagement between drops due to the shape of stabilizer particles and with an insignificant participation of background electrolyte cations in the formation of interdroplet interactions. Therefore, the resulting aggregates are unstable and easily destroyed, even under weak mechanical stress. As a consequence, the acid-base properties of the pCNC surface are practically independent of the emulsion preparation method (with or without electrolyte) as well as the concentration of the background electrolyte. The reduced viscosity of pCNC-stabilized emulsions in the presence of an electrolyte, coupled with the absence of acute toxicity, allows us to recommend them as a convenient oral delivery system for fat-soluble, biologically active substances. Our emulsions carrying donepezil (an anti-Alzheimer drug) showed better performance than a solution of donepezil hydrochloride in preventing memory impairment tested on laboratory mice.

Classification of l-DOPA pharmacokinetics shapes and creating a predictive model

BackgroundIt is known that the pharmacokinetics (PK) of levodopa (LD) varies considerably. Difference in PK shapes is expected to affect drug efficacy and development of dyskinesia. In this study, the authors aimed to explore correlations between PK series data of LD and clinical characteristics and dyskinesia in patients with Parkinson's disease (PD). MethodsWe studied 270 PD patients who underwent PK assessment after administration of LD/carbidopa (100/10 mg) in non-compartmental analysis. The patients were grouped according to similarities in time series data of blood LD concentration. Each group was analyzed with respect to clinical characteristics and PK parameters. We created a model to predict PK patterns based on these findings. ResultsPD patients were divided into three groups by clustering analysis: blood LD concentration of the patients in Groups 1 (n = 129), 3 (n = 44) and 2 (n = 97) rose rapidly, relatively slowly and at an intermediate rate, respectively. There were no statistically significant differences in patient characteristics except age among the three groups (one-way ANOVA). Multivariate analysis showed that frequency of dyskinesias in Group 1 was significantly higher than that in Group 2. We successfully created a PK pattern prediction model based on body weight and blood LD concentration at 15 and 30 min after administration. ConclusionsThe PK series data of LD was classified into three patterns. The rapid absorption was associated with dyskinesias. Patients’ PK patterns were successfully predicted based on their body weight and two-point LD concentration.

Hole-driven dynamics of a three-dimensional gravitational liquid curtain

It is known that the disintegration of vertical liquid curtains (sheets) is affected crucially by the amplification of free edge holes forming inside the curtain. This paper aims to investigate the influence of the hole expansion dynamics, driven by the so-called rim retraction, on the breakup of a liquid curtain, in both supercritical (Weber number $We > 1$ ) and subcritical ( $We < 1$ ) conditions. The analysis is based on three-dimensional direct numerical simulations. For a selected supercritical configuration, the steady flow topology is first analysed. The investigation reveals the classic triangular shape regime of the steady curtain, due to the surface-tension-induced borders retraction towards its centre plane. The unsteady dynamics is then investigated as the curtain response to a hole perturbation introduced artificially in the steady flow configuration. The hole evolution determines a rim retraction phenomenon inside the curtain, which is influenced by both capillary and gravity forces. In supercritical conditions, the hole does not influence the curtain flow dynamics in the long-time limit. By reducing the Weber number slightly under the critical threshold ( $We=1$ ), the initial amplification rate of the hole area increases, due to the stronger retraction effect of surface tension acting on the hole rims. The free hole expansion in fully subcritical conditions ( $We < 1$ ) is investigated finally by simulating an edge-free curtain flow. As $We$ decreases progressively, the hole expands while it is convected downstream by gravity acceleration. In the range $0.4< We<1$ , the subcritical curtain returns to the intact unperturbed configuration after the hole expulsion at the downstream outflow. For $We<0.4$ , the surface tension force becomes strong enough to reverse the gravitational motion of the hole top point, which retracts upstream towards the sheet inlet section while expanding along the lateral directions. This last phenomenon causes finally the breakup of the curtain, which results in a columnar regime strictly resembling similar experimental findings of the literature.

Starch granule size and shape characterization of yam (Dioscorea alata L.) flour using automated image analysis.

Roots, tubers and bananas (RTB) play an essential role as staple foods, particularly in Africa. Consumer acceptance for RTB products relies strongly on the functional properties of, which may be affected by the size and shape of its granules. Classically, these are characterized either using manual measurements on microscopic photographs of starch colored with iodine, or using a laser light-scattering granulometer (LLSG). While the former is tedious and only allows the analysis of a small number of granules, the latter only provides limited information on the shape of the starch granule. In this study, an open-source solution was developed allowing the automated measurement of the characteristic parameters of the size and shape of yam starch granules by applying thresholding and object identification on microscopic photographs. A random forest (RF) model was used to predict the starch granule shape class. This analysis pipeline was successfully applied to a yam diversity panel of 47 genotypes, leading to the characterization of more than 205 000 starch granules. Comparison between the classical and automated method shows a very strong correlation (R2 = 0.99) and an absence of bias for granule size. The RF model predicted shape class with an accuracy of 83%. With heritability equal to 0.85, the median projected area of the granules varied from 381 to 1115 μm2 and their observed shapes were ellipsoidal, polyhedral, round and triangular. The results obtained in this study show that the proposed open-source pipeline offers an accurate, robust and discriminating solution for medium-throughput phenotyping of yam starch granule size distribution and shape classification. © 2023 The Authors. Journal of The Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Shape Classification of Cloud Particles Recorded by the 2D-S Imaging Probe Using a Convolutional Neural Network

Shape Classification of Cloud Particles Recorded by the 2D-S Imaging Probe Using a Convolutional Neural Network

PatchMixer: Rethinking network design to boost generalization for 3D point cloud understanding

The recent trend in deep learning methods for 3D point cloud understanding is to propose increasingly sophisticated architectures either to better capture 3D geometries or by introducing possibly undesired inductive biases. Moreover, prior works introducing novel architectures compared their performance on the same domain, devoting less attention to their generalization to other domains. We argue that the ability of a model to transfer the learnt knowledge to different domains is an important feature that should be evaluated to exhaustively assess the quality of a deep network architecture. In this work we propose PatchMixer, a simple yet effective architecture that extends the ideas behind the recent MLP-Mixer paper to 3D point clouds. The novelties of our approach are the processing of local patches instead of the whole shape to promote robustness to partial point clouds, and the aggregation of patch-wise features using an MLP as a simpler alternative to the graph convolutions or the attention mechanisms that are used in prior works. We evaluated our method on the shape classification and part segmentation tasks, achieving superior generalization performance compared to a selection of the most relevant deep architectures.

Galaxy Maps: Localized Foliations for Bijective Volumetric Mapping

A method is presented to compute volumetric maps and parametrizations of objects over 3D domains. As a key feature, continuity and bijectivity are ensured by construction. Arbitrary objects of ball topology, represented as tetrahedral meshes, are supported. Arbitrary convex as well as star-shaped domains are supported. Full control over the boundary mapping is provided. The method is based on the technique of simplicial foliations, generalized to a broader class of domain shapes and applied adaptively in a novel localized manner. This increases flexibility as well as efficiency over the state of the art, while maintaining reliability in guaranteeing map bijectivity.

Deployable strip structures

We introduce the new concept of C-mesh to capture kinetic structures that can be deployed from a collapsed state. Quadrilateral C-meshes enjoy rich geometry and surprising relations with differential geometry: A structure that collapses onto a flat and straight strip corresponds to a Chebyshev net of curves on a surface of constant Gaussian curvature, while structures collapsing onto a circular strip follow surfaces which enjoy the linear-Weingarten property. Interestingly, allowing more general collapses actually leads to a smaller class of shapes. Hexagonal C-meshes have more degrees of freedom, but a local analysis suggests that there is no such direct relation to smooth surfaces. Besides theory, this paper provides tools for exploring the shape space of C-meshes and for their design. We also present an application for freeform architectural skins, namely paneling with spherical panels of constant radius, which is an important fabrication-related constraint.