liveBook · Manning

MIDV-500

As explained in Machine Learning Platform Engineering

MIDV-500 is a dataset of video clips depicting multiple types of identity documents for document analysis and detection tasks.

MIDV-500 contains 500 video clips spanning 50 identity-document types, including ID cards, passports, and driving licences. For an ID-card object-detection workflow, filter the dataset to ID cards before preparing images and labels. section 7.3.1

A reproducible preparation pipeline separates three jobs:

  • Download and extract the data in a Kubeflow component. section 7.3.1
  • Split paired images and labels into training, validation, and test outputs; preserve their matching filenames. section 7.3.1
  • Inspect output files as a sanity check before using the splits. section 7.3.1
Listing 7.9 Configuration training YOLOv8 with a custom dataset
path: /home/jovyan/data/
train: "train/images"
val: "val/images"
test: "test/images"
names:
  0: id_card

The YOLOv8 configuration points to the prepared train, validation, and test image directories and names the detection class `id_card`. Keep the corresponding label directories alongside the images. section 7.3.1

How Machine Learning Platform Engineering covers it

How deep it goes
Defines the dataset’s size and document types, then develops download, splitting, and inspection steps rather than dataset collection methods.
Examples
Uses an ID-card subset, paired image and label files, and train, validation, and test directories.
What it assumes
Introduces Kubeflow component creation from a Python function while using Python data-handling code.
What you can do after
Apply it via Kubeflow Pipelines to download, split, and inspect ID-card data for YOLOv8.
How it connects
Links MIDV-500 to object detection, Kubeflow Pipelines, YOLOv8, dataset splitting, and MinIO.
Cover of Machine Learning Platform Engineering
liveBookmanning.com
Read Machine Learning Platform Engineering on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Machine Learning Platform Engineering. The section links point to the parts of the book it draws on.