liveBook · Manning

Vision Transformer

As explained in Fight Fraud with Machine Learning

A vision transformer is a transformer-based neural-network model adapted to process visual data such as images.

A vision transformer (ViT) adapts transformer-based modeling from textual data to visual data. In fraud detection, it supports visual tasks such as extracting information from identity documents for automated know-your-customer (KYC) verification. section 4.4 section 4.3.4

Use vision transformers to:

  • Process visual identity documents for automated KYC workflows. section 4.3.4
  • Extend transformer-based attention modeling beyond text to visual fraud-detection data. section 4.4
Figure 4.19 Transformer model architecture containing multi-head attention layers that use attention mechanism only to extract complex information from input sequence.

The transformer architecture figure provides context for the multi-head attention mechanism underlying the model family from which vision transformers emerged. section 4.3.4

How Fight Fraud with Machine Learning covers it

How deep it goes
Defines the visual adaptation and application, but stops short of explaining ViT architecture, training, or implementation.
Examples
Uses identity documents as the visual artifact for extracting information during KYC verification.
What it assumes
Builds on an understanding of transformers and attention while treating visual adaptation as the central extension.
What you can do after
Apply it to automate KYC identity verification from visual identity documents.
How it connects
Connects transformers, attention mechanisms, visual data, KYC, and identity verification.

Also explained in

Compare how 8 books explain Vision Transformer →

Related concepts

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