README.md

Vision Transformer

This repository contains PyTorch implementation of original paper: An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

"While the Transformer architecture has become the de-facto standard for natural language processing tasks, its applications to computer vision remain limited. In vision, attention is either applied in conjunction with convolutional networks, or used to replace certain components of convolutional networks while keeping their overall structure in place. We show that this reliance on CNNs is not necessary and a pure transformer applied directly to sequences of image patches can perform very well on image classification tasks. When pre-trained on large amounts of data and transferred to multiple mid-sized or small image recognition benchmarks (ImageNet, CIFAR-100, VTAB, etc.), Vision Transformer ( ViT) attains excellent results compared to state-of-the-art convolutional networks while requiring substantially fewer computational resources to train."

Installation

HTTPS:

$ pip install git+https://github.com/Danielto1404/vision-transformer.git

SSH:

$ pip install [email protected]:Danielto1404/vision-transformer.git

VIT Feature extractor example

import torch
from vit import ViT

model = ViT(
    image_size=(3, 28, 28),  # channels x height x width  
    patch_size=14,  # n x n patch
    embedding_dim=768,  # embedding dimension which
    layers=4,  # number of transformer encoder layers
    heads=12,  # number of transformer encoder heads
    head_dim=64,  # single head dimension
    feedforward_dim=2048,  # transformer encoder mlp dimension
    dropout=0.2,  # dropout
    pooling="cls"  # [`cls`, `mean`]
)

x = torch.rand(32, 3, 28, 28)  # batch x channels x height x width
features = model(x)  # batch x embedding_dim

Classifier Example

import torch
from vit import ViTForClassification

model = ViTForClassification(
    num_classes=10,  # number of classes
    image_size=(3, 28, 28),  # channels x height x width  
    patch_size=14,  # n x n patch
    embedding_dim=768,  # embedding dimension
    layers=4,  # number of transformer encoder layers
    heads=12,  # number of transformer encoder heads
    head_dim=64,  # single head dimension
    feedforward_dim=2048,  # transformer encoder mlp dimension
    dropout=0.2,  # dropout
    pooling="cls"  # [`cls`, `mean`]
)

x = torch.rand(32, 3, 28, 28)  # batch x channels x height x width
classes = model(x)  # batch x num_classes (32 x 10)

Pretrained models

There are a few pretrained models: TODO