3D Model Retrieval Based on a 3D Shape Knowledge Graph

Abstract

A development of 3D construction technology has led to 3D models being applied in many fields, and the number of 3D models has exploded in recent years. Thus, 3D model retrieval has become a popular topic with many proposed approaches. However, all of the methods focus on the 3D model's global structural descriptor design based on various deep learning networks and ignore the local structural information of the 3D model and the correlation of the local structures. In this paper, we propose a novel 3D model retrieval method based on a 3D shape knowledge graph. We first introduce the concept of a geometric word that can be utilized to assemble other 3D model. Second, we construct a 3D shape knowledge graph based on the geometric words, models and their relations. Additionally, we propose a novel graph embedding method to generate embeddings of nodes. Finally, an effective multiple entities' retrieval method is used to handle the 3D model retrieval problem. More specifically, the 3D shape knowledge graph retains the basic structural information and saves these as a set of triples. Any 3D model can find its geometric words in a rich enough knowledge graph. It is reasonable that our approach can solve the cross-domain model retrieval problem. Our approach focuses on the structural information of 3D model and is not restricted by the database. We evaluate the proposed method on the ModelNet40 dataset for the 3D model retrieval task. Meanwhile, we also utilize the ShapeNet dataset to evaluate the performance of cross-domain retrieval task. Experimental results and comparisons with state-of-the-art methods demonstrate that our framework can achieve superior performance.

CAD 3D Model classification by Graph Neural Networks: A new approach based on STEP format

“…In [QRS*16] and [QYSG17], Qi et al directly used point cloud data, represented by three coordinates to feed a network able to perform feature transformations and aggregate data points by max pooling. In [NWSL20], authors utilized the PointNet++ model to segment each 3D model into a set of shapes and extracted their features. The K-means method was then utilized to construct from each different shape a node of a 3D shape knowledge graph.…”

3D-CSD+: Extração de características 3D baseada em grafos

“…A classificação de objetos tridimensionais, necessária para a percepção do ambiente, normalmente se baseia em encontrar uma correspondência entre duas formas baseada em alguma métrica de similaridade, normalmente a similaridade visual, ou seja, duas formas que são visualmente parecidas (BUSTOS; SIPIRAN, 2012). Esse processo normalmente inclui as seguintes etapas (NIE et al, 2020):…”

“…Ainda, sistemas autônomos críticos possuem um risco muito maior do que sistemas simulados -um erro em um robô ou um veículo autônomo pode causar danos permanentes ao próprio sistema, à propriedade de terceiros ou atingir pessoas. É extremamente necessário que o sistema seja robusto e que forneça respostas em tempo real para aumentar a segurança (NIE et al, 2020). Em muitos estudos, não é informada a taxa de atualização necessária para a detecção/classificação do objeto, ou mostram taxas muito distantes do ideal para respostas em tempo real.…”