Vision transformer
lastmod 2025-11-12created 2024-02-066 edits
A vision transformer (ViT) is a transformer designed for computer vision. A ViT breaks down an input image into a series of patches (rather than breaking up text into tokens), serialises each patch into a vector, and maps it to a smaller dimension with a single matrix multiplication. These vector embeddings are then processed by a transformer encoder as if they were token embeddings.
ViT has found applications in image recognition, image segmentation, and autonomous driving.
A vision transformer is an image processing model built with a transformer. It started from the idea, introduced in the 2020 paper "An image is worth 16x16 words", of processing an image as if it were words, and depending on the field it achieves performance similar to conventional convolutional neural networks.
Applies the Transformer, widely used in natural language (NLP), to Vision Tasks.
Categories
See also
Favorite site
- Vision transformer - Wikipedia
- Transformer-based image segmentation - Hugging Face Community Computer Vision Course
Paper reviews/notes/summaries
- (Paper summary) Transformer finally arrives in the Vision field - ViT : Vision Transformer(2020)
- Vision Transformer Explained | Papers With Code
- Vision Transformer (ViT)
- Vision Transformer(ViT) paper review
- Vision Transformer (ViT) | Standing-O
- Vision transformer made easy .. : Naver Blog
- Vision Transformer(ViT) paper review
- What are the Transformer and the Attention Mechanism?
- (AI/ViT) Vision Transformer(ViT), understanding it easily with pictures