Dataset · Andyson
Industrial defect benchmark with semantic annotation
Andyson/DefectSpectrum
A collection of manufacturing defect images with segmentation masks and descriptions for quality control systems.
What's inside
The benchmark combines 1,000 to 10,000 industrial defect images with detailed English semantic annotations. It is suitable for image segmentation and image caption generation tasks. The card notes that the full version of the dataset has been moved to a separate repository.
What you can do with it
- Training production defect segmentation models
- Testing VLM models on recognizing and describing defects from photos
- Tasks
- Image segmentation, Image → text
- Languages
- English
- Size
- 1K–10K rows
- Downloads this month
- 19.8K
- Likes
- 3
How to load it
from datasets import load_dataset
ds = load_dataset("Andyson/DefectSpectrum")