Object detection model

As explained in Machine Learning Platform Engineering

An object detector is a machine learning model that identifies object instances and estimates their locations in visual input.

An object detection model identifies object instances and estimates their locations in visual input. In a production identity-card workflow, it turns a labeled image dataset into a detector that can recognize classes absent from its original training data. section 2.1.3

A practical detector workflow includes:

• Start with a pretrained YOLO model, whose COCO-trained classes do not include identity cards or passports. section 2.1.3
• Fine-tune YOLOv8 using a well-labeled identity-card dataset. section 2.1.3
• Separate images for training, testing, and validation, then use the training library's evaluation logic. section 2.1.3

Identity-card detection

A COCO-trained YOLO model cannot detect identity cards out of the box. Fine-tuning YOLOv8 on labeled identity-card images adapts the detector to that target class. section 2.1.3

Figure 2.5 A view of a retraining pipeline using the modular codebase concept. This approach of keeping the model, code, configuration files, and data as distinct versioned components with lineage links ensures that the process remains flexible, fast, and adaptable while enabling experimentation, debugging, and iterative development.

A modular retraining pipeline keeps the model, code, configuration, and data as distinct versioned components linked by lineage, supporting experimentation, debugging, and iterative development. section 2.1.3

How Machine Learning Platform Engineering covers it

How deep it goes

Explains pretrained-model adaptation, dataset splitting, and library-based evaluation, but does not detail detector internals.

Examples

Uses identity-card detection, COCO classes, YOLOv8 fine-tuning, and separate training, testing, and validation images.

What it assumes

Assumes labeled images and introduces adaptation from COCO-trained classes to a domain-specific detection target.

What you can do after

Build a custom identity-card detector by fine-tuning YOLOv8 on labeled images and evaluating held-out data.

How it connects

Connects object detection with YOLO, fine-tuning, labeled datasets, model evaluation, and retraining pipelines.

liveBookmanning.com

Read Machine Learning Platform Engineering on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Math and Architectures of Deep LearningA model that predicts object classes and their spatial locations in an image.

Compare how 2 books explain object detection model →

Related concepts

• YOLOincludes · YOLO is a family of single-stage object-detection models that predict object classes and locations directly from an image in one inference pass.
• RetinaNetincludes · RetinaNet is a one-stage object-detection architecture that uses convolutional feature pyramids with separate classification and bounding-box regression heads.
• object detectionused by · Object detection is a computer-vision task that identifies object instances in visual data, assigns them categories, and estimates their locations.
• computer visionused by · Computer vision is the field of computing concerned with using computational methods to extract and interpret information from images, video, and other visual data.
• single-stage detectorincludes · An object-detection architecture that predicts object locations and class labels directly in one integrated model pass without a separate region-proposal stage.

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.