MVIP
收藏资源简介:
MVIP是一个面向应用的多视角和多模态工业零件识别数据集,由弗劳恩霍夫IPK研究所创建。该数据集包含了校准过的RGBD多视角图像以及对象的物理属性、自然语言描述和超类别等信息。数据集共包含约570,000张图像,分为训练集、验证集和测试集,适用于工业零件识别相关的研究,旨在解决小样本学习、视觉相似零件识别等问题。
MVIP is an application-oriented multi-view and multi-modal industrial part recognition dataset developed by the Fraunhofer IPK Institute. This dataset contains calibrated RGBD multi-view images, along with information including the physical properties of the objects, natural language descriptions, and super-categories. In total, the dataset includes approximately 570,000 images, which are divided into training, validation, and test sets. It is suitable for research related to industrial part recognition, aiming to address issues such as few-shot learning and the recognition of visually similar industrial parts.

- 1MVIP -- A Dataset and Methods for Application Oriented Multi-View and Multi-Modal Industrial Part Recognition弗劳恩霍夫IPK研究所 · 2025年



