Single-stage detector

As explained in Deep Learning with Python, Third Edition

An object-detection architecture that predicts bounding boxes and class labels in one pass with a single model, rather than using separate proposal and classification stages.

A single-stage detector finds objects and their locations in one pass: one model jointly predicts bounding boxes, labels, and confidence scores instead of first proposing regions and then classifying them. This design favors speed and efficiency, making it useful for real-time or resource-constrained applications. section 12.1.2 section 12.1.1

Figure 12.1 Object detectors draw boxes around objects in an image and label them.

Bounding boxes show the detector’s practical output: each detected object receives a location and a label. section 12.0

A simplified YOLO detector

YOLO divides a 448 × 448 image into a grid and predicts one bounding box, confidence score, and class label at each grid location. A ConvNet backbone extracts features; a prediction head reshapes them into grid outputs, separating five box-related values from class probabilities. Empty grid locations should produce near-zero confidence. section 12.2.2

Use the architecture when:

• You need fast detection for real-time applications such as self-driving systems. section 12.4
• Bounding boxes are sufficient, because segmentation would add pixel-level labeling and computational costs. section 12.0
• You accept a possible minor accuracy tradeoff for greater speed and efficiency than two-stage detectors. section 12.1.2

How Deep Learning with Python, Third Edition covers it

How deep it goes

Explains joint outputs, grid prediction, confidence, and backbones through a simplified YOLO build, without developing the full original architecture.

Examples

Uses counting, tracking, and cropping applications plus a grid-based YOLO model with ResNet and KerasHub.

What it assumes

Assumes familiarity with ConvNet backbones while introducing object detection through bounding boxes, labels, grids, and confidence.

What you can do after

Build a simplified YOLO detector from scratch using a pretrained ConvNet backbone and separate box and class outputs.

How it connects

Connects object detection with bounding boxes, ConvNets, YOLO, RetinaNet, SSD, R-CNN, image segmentation, and feature pyramid networks.

liveBookmanning.com

Read Deep Learning with Python, Third Edition on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Deep Learning with R, Third EditionA single-stage object detector predicts object locations and categories directly in one integrated model pass rather than first generating region proposals.

Compare how 2 books explain single-stage detector →

Related concepts

• RetinaNetincludes · RetinaNet is a one-stage object-detection architecture that uses convolutional feature pyramids with separate classification and bounding-box regression heads.
• 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.
• object detection modelkind of · An object detection model is a computer-vision model that identifies object instances and predicts their classes and spatial locations in visual data.
• object detectionaddresses · Object detection is a computer-vision task that identifies object instances in visual data, assigns them categories, and estimates their locations.

This explainer was generated by Manning's knowledge graph from the text of Deep Learning with Python, Third Edition. The section links point to the parts of the book it draws on.