Vision Transformers (ViT) — Roadmap Step & Resources
Applying the transformer architecture to images by treating patches as tokens
- Level: advanced
- Estimated time: 1 days
- Roadmap: Deep Learning & Neural Networks
Before this step
Study resources
- An Image is Worth 16x16 Words (ViT) (Article) — The original Vision Transformer paper applying Transformers to image classification.
- Vision Transformers Explained (Video) — A practical explanation of applying self-attention to image patches.
Part of
- Transformers & Modern Architectures — section
- Deep Learning & Neural Networks roadmap — the full learning path