Skip to content

Dataset · pmchard

3D Additive Manufacturing Defect Dataset

pmchard/3D-ADAM

Raw data, images, and English text for anomaly detection and segmentation in 3D printing.

What's inside

The 3D-ADAM (3D Anomaly Detection in Additive Manufacturing) dataset contains 1 to 10 thousand records with images and text data in English. The repository contains raw data of the full set, separated by individual parts (part-instance). Separately, the authors mention a subset from the MechMind-Nano camera for unsupervised learning via the anomalib library.

What you can do with it

  • Training quality control and defect detection models in 3D printing
  • Testing algorithms for segmentation and localization of manufacturing defects
Languages
English
Size
1K–10K rows
Downloads this month
115.9K
Likes
11

How to load it

from datasets import load_dataset

ds = load_dataset("pmchard/3D-ADAM")