Layer-wise Imaging Dataset from Powder Bed Additive Manufacturing Processes for Machine Learning Applications (Peregrine v2022-10)

This release consists of six datasets which together include multi-modal layer-wise powder bed images from two different powder bed printing technologies. These datasets are designed primarily to facilitate the development and testing of new computer vision and machine learning based anomaly and defect detection algorithms. The authors provide both training data with corresponding ground truth pixel masks and evaluation data with corresponding baseline prediction pixel masks made by a trained neural network. The laser powder bed fusion (L-PBF) datasets are sourced from EOS M290 and AddUp FormUp 350 printers and the binder jet (BJ) dataset is sourced from an ExOne M-Flex printer. The materials represented in these datasets include 17-4 PH Stainless Steel, DMREF, Inconel 718, Maraging Steel, and H13 Steel. The sensor imaging modalities represented include visible-light (VL), temporally-integrated (i.e., long duration exposure) near-infrared (TI-NIR), and wide-band infrared (IR). To download the dataset: 1. Create a Globus account. 2. Create a Globus Endpoint on your computer. 3. Transfer the dataset from the OLCF DOI-DOWNLOADS Collection to your Collection. Common troubleshooting steps: 1. Confirm that the transfer is going from OLCF DOI-DOWNLOADS to your Collection. 2. Create an exception for Globus in your antivirus software so that it can create an Endpoint. 3. Manually create a Globus access directory (where the data will be downloaded) by going to the Preferences > Access tab.

搜集汇总

背景与挑战

背景概述

该数据集是一个用于机器学习应用的粉末床增材制造分层成像数据集，包含六个子集，涵盖激光粉末床融合和粘结剂喷射两种技术，提供多模态图像以支持异常检测算法的开发。数据包括带真实掩码的训练集和带基线预测的评估集，材料涉及多种钢材和合金，成像模态包括可见光、近红外和红外。数据集设计旨在促进计算机视觉和机器学习在增材制造质量控制中的研究和应用。