CADSketchNet

CADSketchNet是一个专门为3D CAD模型检索系统设计的标注草图数据集，由印度理工学院马德拉斯分校和国家理工学院蒂鲁奇拉帕利创建。该数据集包含59,497个草图，分为两部分：Dataset-A包含58,696个计算机生成的草图，Dataset-B包含801个手绘草图。数据集的创建涉及从现有CAD数据库收集3D模型，并生成草图数据，包括计算机生成的和手绘的草图。CADSketchNet旨在支持开发基于深度学习的检索系统，该系统能够通过草图图像作为输入查询来检索3D CAD模型。数据集的应用领域包括工程设计和产品开发，旨在通过草图快速检索和重用现有的3D CAD模型，提高设计效率和创新。

CADSketchNet is an annotated sketch dataset specifically designed for 3D CAD model retrieval systems, created by the Indian Institute of Technology Madras and National Institute of Technology Tiruchirappalli. This dataset contains 59,497 sketches, divided into two subsets: Dataset-A consists of 58,696 computer-generated sketches, and Dataset-B includes 801 hand-drawn sketches. The development of CADSketchNet involves collecting 3D models from existing CAD databases and generating sketch data covering both computer-generated and hand-drawn sketches. CADSketchNet aims to support the development of deep learning-based retrieval systems that can retrieve 3D CAD models using sketch images as input queries. Its application fields include engineering design and product development, with the goal of quickly retrieving and reusing existing 3D CAD models via sketches to improve design efficiency and foster innovation.

搜集汇总

背景与挑战

背景概述

CADSketchNet是一个用于3D CAD模型检索的标注草图数据集，包含两个子集：Dataset-A有58,696个计算机生成草图（68个类别），Dataset-B有801个手绘草图（42个类别）。该数据集旨在支持基于深度神经网络的3D模型检索研究，采用CC BY-NC-SA许可证，限制商业使用。