3D Model Classification Research Articles

MLGCN: an ultra efficient graph convolutional neural model for 3D point cloud analysis.

With the rapid advancement of 3D acquisition technologies, 3D sensors such as LiDARs, 3D scanners, and RGB-D cameras have become increasingly accessible and cost-effective. These sensors generate 3D point cloud data that require efficient algorithms for tasks such as 3D model classification and segmentation. While deep learning techniques have proven effective in these areas, existing models often rely on complex architectures, leading to high computational costs that are impractical for real-time applications like augmented reality and robotics. In this work, we propose the Multi-level Graph Convolutional Neural Network (MLGCN), an ultra-efficient model for 3D point cloud analysis. The MLGCN model utilizes shallow Graph Neural Network (GNN) blocks to extract features at various spatial locality levels, leveraging precomputed KNN graphs shared across GCN blocks. This approach significantly reduces computational overhead and memory usage, making the model well-suited for deployment on low-memory and low-CPU devices. Despite its efficiency, MLGCN achieves competitive performance in object classification and part segmentation tasks, demonstrating results comparable to state-of-the-art models while requiring up to a thousand times fewer floating-point operations and significantly less storage. The contributions of this paper include the introduction of a lightweight, multi-branch graph-based network for 3D shape analysis, the demonstration of the model's efficiency in both computation and storage, and a thorough theoretical and experimental evaluation of the model's performance. We also conduct ablation studies to assess the impact of different branches within the model, providing valuable insights into the role of specific components.

Multi-view Consistent and Complementary Information Fusion Method for 3D Model Classification

Multi-view Consistent and Complementary Information Fusion Method for 3D Model Classification

QSCC: A Quaternion Semantic Cell Convolution Graph Neural Network for MBD Product Model Classification

Model-based Definition (MBD) 3D engineering has gained wide popularity which enables smart manufacturing. Its tip orientation to the effective reuse of MBD-based product models for creating efficient digital twins. The deep learning-based 3D model classification is an important way in the most recent years, but the existing methods only concern geometric information with very little or no manufacturing information. In this paper, we proposed quaternion semantic cell convolution (QSCC), the quaternion semantic cell convolution graph neural network for MBD product model classification. It is mainly composed of the quaternion product semantic cell(QPSC) and the quaternion simple graph convolution(QSGC). The QPSC is established to fully express the digital twins' semantic information, which includes the multi-dimensional PMI information of the MBD model. The QSGC is established by embedding QPSC into the quaternion vector space, which serializes the multi-granularity semantic features to learn the context semantics through GRU. Compared with existing methods, our approach can solve the problem of model classification with similar geometry but different functions. Specifically, our approach can solve the ambiguity problem caused by blend features such as chamfer. We give quantitative and qualitative evaluations to demonstrate the accuracy and efficiency.

3D Model classification based on regnet design space and voting algorithm

3D Model classification based on regnet design space and voting algorithm

FuS-GCN: Efficient B-rep based graph convolutional networks for 3D-CAD model classification and retrieval

Performing 3-dimensional computer-aided design (3D-CAD) model classification, retrieval, and reuse is of vital importance in industrial manufacturing, as it considerably shortens the engineering development cycle and reduces development costs. Although existing 3D model classification and retrieval methods achieve satisfactory performance when operating on meshes or point clouds, they cannot be applied directly to 3D-CAD models, which are generally represented by the boundary representation (B-rep). To address this issue, and to fully exploit the topology of B-rep, a graph structure descriptor called B-rep graph is proposed to pre-process B-rep data, and to extract the precise topological and geometric features from 3D-CAD models. Meanwhile, a novel efficient neural network called FuS-GCN, based on graph convolutional networks (GCNs), is designed to handle this graph data. To better extract the graph features and to improve the effect of pooling, the self-attention mechanism and feature fusion are incorporated into the pooling layer, yielding the proposed fusion self-attention graph pooling (FuSPool) algorithm. Finally, we demonstrate the effectiveness of FuS-GCN on 3D-CAD model data, while outperforming alternative 3D shape descriptors such as point clouds, voxels, and meshes.

Combine EfficientNet and CNN for 3D model classification.

With the development of multimedia technology, the number of 3D models on the web or in databases is becoming increasingly larger and larger. It becomes more and more important to classify and retrieve 3D models. 3D model classification plays important roles in the mechanical design field, education field, medicine field and so on. Due to the 3D model's complexity and irregularity, it is difficult to classify 3D model correctly. Many methods of 3D model classification pay attention to local features from 2D views and neglect the 3D model's contour information, which cannot express it better. So, accuracy the of 3D model classification is poor. In order to improve the accuracy of 3D model classification, this paper proposes a method based on EfficientNet and Convolutional Neural Network (CNN) to classify 3D models, in which view feature and shape feature are used. The 3D model is projected into 2D views from different angles. EfficientNet is used to extract view feature from 2D views. Shape descriptors D1, D2, D3, Zernike moment and Fourier descriptors of 2D views are adopted to describe the 3D model and CNN is applied to extract shape feature. The view feature and shape feature are combined as discriminative features. Then, the softmax function is used to determine the 3D model's category. Experiments are conducted on ModelNet 10 dataset. Experimental results show that the proposed method achieves better than other methods.

Multi-view SoftPool attention convolutional networks for 3D model classification.

Existing multi-view-based 3D model classification methods have the problems of insufficient view refinement feature extraction and poor generalization ability of the network model, which makes it difficult to further improve the classification accuracy. To this end, this paper proposes a multi-view SoftPool attention convolutional network for 3D model classification tasks. This method extracts multi-view features through ResNest and adaptive pooling modules, and the extracted features can better represent 3D models. Then, the results of the multi-view feature extraction processed using SoftPool are used as the Query for the self-attentive calculation, which enables the subsequent refinement extraction. We then input the attention scores calculated by Query and Key in the self-attention calculation into the mobile inverted bottleneck convolution, which effectively improves the generalization of the network model. Based on our proposed method, a compact 3D global descriptor is finally generated, achieving a high-accuracy 3D model classification performance. Experimental results showed that our method achieves 96.96% OA and 95.68% AA on ModelNet40 and 98.57% OA and 98.42% AA on ModelNet10. Compared with a multitude of popular methods, our algorithm model achieves the state-of-the-art classification accuracy.

Fine-Grained 3D Model Classification Based on Deep Ensemble and Detail Awareness

Deep learning based 3D model classification methods have poor effectiveness in fine-grained 3D model classification. Aiming at the problem, an end-to-end fine-grained 3D model classification framework is proposed, and a network based on deep ensemble learning network and context detail awareness module (CDAM) is constructed. Inputting the multiply views of a 3D model, the global shape features are extracted through the deep ensemble learning sub-network. And the local detail features are obtained through the auxiliary sub-network based on CDAM. Based on above two sub-networks, an end-to-end weakly supervision fine-grained 3D model classification network is constructed. Experiments are conducted on three sub-datasets with different levels of difficulties, Airplane, Chair and Car, from the public dataset FG3D. The classification accuracies for above three sub-datasets are 96.31%, 85.44% and 79.62% respectively, which demonstrate the fine classification performance and more generalization of the proposed method.

Semantically guided projection for zero-shot 3D model classification and retrieval

Semantically guided projection for zero-shot 3D model classification and retrieval

ZS3D-Net: Zero-Shot Classification Network for 3D Models

Zero-shot 3D model classification is very important for the understanding and analysis of 3D models. Aiming at the problems of lack of corresponding datasets and low accuracy of zero-shot 3D model classification, a 3D model dataset ZS3D is constructed and a deep learning network ZS3D-Net is proposed. The dataset consists of 41 classes, 1 677 non-rigid 3D models with complete attributes of all classes, which can be regarded as the benchmark for zero-shot 3D model classification task. For the network, firstly, the visual features of the 3D models are effectively extracted through an ensemble learning sub-network. Then, the correlation between the visual features and semantic features of the unseen and seen classes can be constructed by a semantic manifold embedding sub-network. Finally, the unseen classes can be recognized based on above two sub-networks. On a traditional 3D model dataset and the proposed ZS3D, ZS3D-Net achieves 30.0% and 58.6% classification accuracy respectively, which are on par or better than the state-of-the-art methods. The experiments also demonstrate that the proposed method has good feasibility and validity.

Joint Local Correlation and Global Contextual Information for Unsupervised 3D Model Retrieval and Classification

Unsupervised 3D model analysis has attracted tremendous attentions with the increasing growth of 3D model data and the extensive human annotations. Many effective methods have been designed to address the 3D model analysis with labeled information, while rare methods devote to unsupervised deep learning due to the difficulty of mining reliable information. In this paper, we propose a novel unsupervised deep learning method named joint local correlation and global contextual information (LCGC) for 3D model retrieval and classification, which mines the reliable triplet set and uses triplet loss to optimize the deep neural network. Our method proposes two schemes: 1) Local self-correlation information learning, which adopts the intra and inter information to construct the view-level triplet set. 2) Global neighbor contextual information learning, which employs the neighbor contextual information to explore the reliable relations among 3D models and construct the model-level triplet set. The above schemes encourage that the selected triple set can been used to improve the discrimination of learned features. Extensive evaluations on two large-scale datasets, ModelNet40 and ShapeNet55, have demonstrated the effectiveness of our proposed method.

3D Model Classification Based on Neural Architecture Search

Differentiable architecture search algorithms are largely suffering from the well-known performance of discretization discrepancy, imbalances-update and aggregation of skip-connections. To weaken this impact, a novel approach named resource balance neural architecture search (RBNAS) is proposed. Firstly, the weak operators that contribute less to performance are gradually cut out by the resource-balanced method. Secondly, to make each operation have its own weight in the search process and contribute to the supernet, the relationship between the operations is changed to cooperation before competition. Finally, Gaussian noise is added into candidate operations to minimize the unfair competition of skip-connections. ModelNet and CIFAR-10 datasets are used to carry out a series of experiments, the results show that the algorithm can minimize discretization discrepancy and avoiding collapse caused by skip-connections. Compared with SGAS, NoisyDARTS, and manually designed network architectures, RBNAS can find a series of Pareto-optimal architectures efficiently and achieve a good performance in point cloud classification.

Classification of 3D CAD Models considering the Knowledge Recognition Algorithm of Convolutional Neural Network

In order to improve the classification effect of the 3D CAD model, this paper combines the knowledge recognition algorithm of convolutional neural network to construct the 3D CAD model classification model. Moreover, this paper analyzes the application of the convolutional neural network algorithm in the classification of 3D CAD models, improves the knowledge recognition algorithm, and builds the basic structure of the convolutional neural network. In addition, this paper builds an intelligent recognition model with the support of the improved algorithm, designs experiments to verify the effect of this paper, and draws a visual display statistical graph based on the statistical test results. Through experimental research, it can be known that the 3D CAD classification model based on the knowledge recognition algorithm of convolutional neural network proposed in this paper can effectively improve the classification effect of the CAD model.

3D Model Classification Based on GCN and SVM

3D model classification is an important task. Now, 3D model is usually expressed as point cloud. Disorder of point cloud brings great difficulty into 3D model classification. In order to classify 3D model correctly, a new classification method combining Graph Convolution Network (GCN) and Support Vector Machine (SVM) is proposed in this paper. Point cloud is sampled. K-Nearest Neighbor ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">KNN</i> ) algorithm is used to find K nearest points of sampling point, and adjacency matrix is established for graph convolution operation. Shape features <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D1, D2, D3</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A3</i> of sampling point are computed based on its K nearest points. Coordinates and shape features of sampling point are combined as discriminative feature. 2-layer graph convolution is used to aggregate disambiguation information of 1-degree and 2-degree adjacent points of sampling point for describing point cloud comprehensively. At the same time, maximum pooling and average pooling are adopted to retain representative information. Finally, SVM is used to classify point clouds. Experimental results show that compared with GCN based on coordinates, the proposed network improves accuracy of 3D model classification by 1.67%. Global and local information can be extracted adequately when 1024 points are sampled from point cloud. When we select 20 nearest points to compute shape features <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D1, D2, D3, A3</i> , local information of point can be described better. Shape features <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D1, D2, D3, A3</i> are combined with coordinates to describe shape and structure of point cloud better. 2-layer graph convolutions are adopted to aggregate information of 1-degree and 2-degree nodes for extracting effective disambiguation features.

LFT-Net: Local Feature Transformer Network for Point Clouds Analysis

6G network enables the rapid connection of autonomous vehicles, the generated internet of vehicles establishes a large-scale point cloud, which requires automatic point cloud analysis to build an intelligent transportation system in terms of the 3D object detection and segmentation. Recently, a great variety of deep convolution networks have been proposed for 3D data analysis, making significant progress in the application of deep learning in 3D computer vision. Inspired by the application of transformer network in 2D computer visual tasks, and in order to increase the expression ability of local fine-grained features, we propose an effective local feature transformer network to learn local feature information and correlations between point clouds. Our network is adaptive to the arrangement of set elements through transformer module, so it is suitable for the feature extraction of local point clouds. In addition, experimental results demonstrate that our LFT-network outperforms the state-of-the-art in 3D model classification tasks on ModelNet40 dataset and segmentation tasks on S3DIS dataset.

3D Model Classification Based on Bayesian Classifier with AdaBoost

With the exponential increasement of 3D models, 3D model classification is crucial to the effective management and retrieval of model database. Feature descriptor has important influence on 3D model classification. Voxel descriptor expresses surface and internal information of 3D model. However, it does not contain topological structure information. Shape distribution descriptor expresses geometry relationship of random points on model surface and has rotation invariance. They can all be used to classify 3D models, but accuracy is low due to insufficient description of 3D model. This paper proposes a 3D model classification algorithm that fuses voxel descriptor and shape distribution descriptor. 3D convolutional neural network (CNN) is used to extract voxel features, and 1D CNN is adopted to extract shape distribution features. AdaBoost algorithm is applied to combine several Bayesian classifiers to get a strong classifier for classifying 3D models. Experiments are conducted on ModelNet10, and results show that accuracy of the proposed method is improved.

Center-push loss for joint view-based 3D model classification and retrieval feature learning

View-based 3D model classification and retrieval are increasingly important in various fields. High classification accuracy and retrieval precision are urgently needed in the related applications. However, these two topics are always considered separately and very few works give an in-depth analysis of their relations. We would like to argue that although the classification and retrieval focus on different characteristics of embedding features, they are compatible rather than opposed to each other. Inspired by the recent deep metric learning approaches in this field, we propose a novelty loss named center-push loss for joint feature learning. The proposed loss can drive the convolutional neural network to learn object features that distributed compact in intra-class while separated in inter-class effectively. It avoids the annoying triplet sampling operation which always needs a delicately designed sampling strategy for efficient network optimization. The new loss is simple in structure and fast in training and achieves the best performance in classification and retrieval compared with many states of the arts.

Integral spin images usage in deep learning algorithms for 3D model classification

We are investigating a problem of 3D model classification using deep learning algorithms. We propose integral spin images usage as 3D model representation. A number of computational experiments were made to build spin images for 3D models of Princeton Shape Benchmark and use them to train LeNet-5, AlexNet and ResNet deep neural networks. The results showed that integral spin images can be used in conjunction with deep learning algorithms in 3D model classification problem. However, with greater number of classes classification accuracy tends to decrease. It is expected that designing a more complex neural network architecture and expanding number of data characteristics can increase accuracy.

3D Mesh Model Classification with a Capsule Network

With the widespread success of deep learning in the two-dimensional field, how to apply deep learning methods from two-dimensional to three-dimensional field has become a current research hotspot. Among them, the polygon mesh structure in the three-dimensional representation as a complex data structure provides an effective shape approximate representation for the three-dimensional object. Although the traditional method can extract the characteristics of the three-dimensional object through the graphical method, it cannot be applied to more complex objects. However, due to the complexity and irregularity of the mesh data, it is difficult to directly apply convolutional neural networks to 3D mesh data processing. Considering this problem, we propose a deep learning method based on a capsule network to effectively classify mesh data. We first design a polynomial convolution template. Through a sliding operation similar to a two-dimensional image convolution window, we directly sample on the grid surface, and use the window sampling surface as the minimum unit of calculation. Because a high-order polynomial can effectively represent a surface, we fit the approximate shape of the surface through the polynomial, use the polynomial parameter as the shape feature of the surface, and add the center point coordinates and normal vector of the surface as the pose feature of the surface. The feature is used as the feature vector of the surface. At the same time, to solve the problem of the introduction of a large number of pooling layers in traditional convolutional neural networks, the capsule network is introduced. For the problem of nonuniform size of the input grid model, the capsule network attitude parameter learning method is improved by sharing the weight of the attitude matrix. The amount of model parameters is reduced, and the training efficiency of the 3D mesh model is further improved. The experiment is compared with the traditional method and the latest two methods on the SHREC15 data set. Compared with the MeshNet and MeshCNN, the average recognition accuracy in the original test set is improved by 3.4% and 2.1%, and the average after fusion of features the accuracy reaches 93.8%. At the same time, under the premise of short training time, this method can also achieve considerable recognition results through experimental verification. The three-dimensional mesh classification method proposed in this paper combines the advantages of graphics and deep learning methods, and effectively improves the classification effect of 3D mesh model.

2DSlicesNet: A 2D Slice-Based Convolutional Neural Network for 3D Object Retrieval and Classification

3D data can be instrumental to the computer vision field as it provides insightful information about the full 3D models' geometry. Recently, with easy access to both computational power and huge 3D databases, it is feasible to apply convolutional neural networks to automatically extract the 3D models' features. This paper presents a novel approach, called 2DSlicesNet, which deals with the issue of 3D model retrieval and classification using a 2D slice-based representation with a 3D convolutional neural network. The assumption in this context is that similar 3D models will be composed of almost identical 2D slices. Therefore, we first transform each normalized 3D model into a set of 2D slices corresponding to its first main axis, and then use them as input data to our 3D convolutional neural network. Experimental results and comparison with state-of-the-art approaches, using ModelNet10 and ModelNet40 datasets, prove that our proposed 2DSlicesNet approach can reach notable rates of accuracy in classification and retrieval.