人工智能培训

搜索

机器学习论文:PARTNER:细粒度和分层零件级3D对象理解的大规模基准(PartNet: A Large-scale Benchmark for Fine-g

[复制链接]
wynrefer 发表于 4 天前 | 显示全部楼层 |阅读模式
wynrefer 4 天前 487 0 显示全部楼层
机器学习论文:PARTNER:细粒度和分层零件级3D对象理解的大规模基准(PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical  Part-level 3D Object Understanding)我们提供了PartNet:一个一致的大型3D对象数据集,用细粒度,实例级和分层3D零件信息进行注释。 Ourdataset包含573,585个零件实例,超过26,671个3D模型,涵盖24个对象类别。该数据集能够并作为许多任务的催化剂,如形状分析,动态三维场景建模和模拟,可供性分析等。使用我们的数据集,我们建立了三个评估三维零件识别的基准测试任务:细粒度语义分割,分层语义分割和实例分割。我们对用于细粒度语义分割的最先进的3D深度学习算法和用于分层语义分割的三种基线方法进行了基准测试。我们还提出了一种新的部分实例分割方法,并展示了优于现有方法的性能。
We present PartNet: a consistent, large-scale dataset of 3D objects annotatedwith fine-grained, instance-level, and hierarchical 3D part information.Ourdataset consists of 573,585 part instances over 26,671 3D models covering 24object categories.This dataset enables and serves as a catalyst for many taskssuch as shape analysis, dynamic 3D scene modeling and simulation, affordanceanalysis, and others.Using our dataset, we establish three benchmarking tasksfor evaluating 3D part recognition: fine-grained semantic segmentation,hierarchical semantic segmentation, and instance segmentation.We benchmarkfour state-of-the-art 3D deep learning algorithms for fine-grained semanticsegmentation and three baseline methods for hierarchical semantic segmentation.We also propose a novel method for part instance segmentation and demonstrateits superior performance over existing methods.机器学习论文:PARTNER:细粒度和分层零件级3D对象理解的大规模基准(PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical  Part-level 3D Object Understanding) R5Q8soJShUQYJOj5.jpg
URL地址:https://arxiv.org/abs/1812.02713     ----pdf下载地址:https://arxiv.org/pdf/1812.02713    ----机器学习论文:PARTNER:细粒度和分层零件级3D对象理解的大规模基准(PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical  Part-level 3D Object Understanding)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

wynrefer当前离线
新手上路

查看:487 | 回复:0

快速回复 返回顶部 返回列表