This repository is the official PyTorch implementation of Global Context Vision Transformers.
Global Context Vision
Transformers
Ali Hatamizadeh,
Hongxu (Danny) Yin, Jan Kautz,
and Pavlo Molchanov.
GC ViT achieves state-of-the-art results across image classification, object detection and semantic segmentation tasks. On ImageNet-1K dataset for classification, the tiny, small and base variants of GC ViT with 28M
, 51M
and 90M
parameters achieve 83.40
, 83.95
and 84.47
Top-1 accuracy, respectively, surpassing comparably-sized prior art such as CNN-based ConvNeXt and ViT-based Swin Transformer by a large margin. Pre-trained GC ViT backbones in downstream tasks of object detection, instance segmentation,
and semantic segmentation using MS COCO and ADE20K datasets outperform prior work consistently, sometimes by large margins.
The architecture of GC ViT is demonstrated in the following:
08/28/2022
- GC ViT and its pre-trained weights are now available as part of timm library.
08/11/2022
- New Pre-trained model weights with improved performance have been released. Please see below for download link.
- GC ViT model has been updated with enhanced global query generator.
06/23/2022
- Pre-trained model weights released. Please see below for download link.
06/17/2022
- GC ViT model, training and validation scripts released for ImageNet-1K classification.
- Pre-trained model checkpoints will be released soon.
GC ViT leverages global context self-attention modules, joint with local self-attention, to effectively yet efficiently model both long and short-range spatial interactions, without the need for expensive operations such as computing attention masks or shifting local windows.
ImageNet-1K Pretrained Models
Name | Acc@1 | Acc@5 | Resolution | #Params | FLOPs | Summary | Download |
---|---|---|---|---|---|---|---|
GC ViT-XXT | 79.80 | 95.09 | 224x224 | 12 | 2.1 | summary | model |
GC ViT-XT | 82.04 | 95.99 | 224x224 | 20 | 2.6 | summary | model |
GC ViT-T | 83.40 | 96.40 | 224x224 | 28 | 4.7 | summary | model |
GC ViT-S | 83.95 | 96.65 | 224x224 | 51 | 8.5 | summary | model |
GC ViT-B | 84.47 | 96.84 | 224x224 | 90 | 14.8 | summary | model |
This repository is compatible with NVIDIA PyTorch docker nvcr>=21.06
which can be obtained in this
link.
The dependencies can be installed by running:
pip install -r requirements.txt
Please download the ImageNet dataset from its official website. The training and validation images need to have sub-folders for each class with the following structure:
imagenet
├── train
│ ├── class1
│ │ ├── img1.jpeg
│ │ ├── img2.jpeg
│ │ └── ...
│ ├── class2
│ │ ├── img3.jpeg
│ │ └── ...
│ └── ...
└── val
├── class1
│ ├── img4.jpeg
│ ├── img5.jpeg
│ └── ...
├── class2
│ ├── img6.jpeg
│ └── ...
└── ...
The GC ViT
model can be trained from scratch on ImageNet-1K dataset by running:
python -m torch.distributed.launch --nproc_per_node <num-of-gpus> --master_port 11223 train.py \
--config <config-file> --data_dir <imagenet-path> --batch-size --amp <batch-size-per-gpu> --tag <run-tag> --model-ema
To resume training from a pre-trained checkpoint:
python -m torch.distributed.launch --nproc_per_node <num-of-gpus> --master_port 11223 train.py \
--resume <checkpoint-path> --config <config-file> --amp --data_dir <imagenet-path> --batch-size <batch-size-per-gpu> --tag <run-tag> --model-ema
To evaluate a pre-trained checkpoint using ImageNet-1K validation set on a single GPU:
python validate.py --model <model-name> --checkpoint <checkpoint-path> --data_dir <imagenet-path> --batch-size <batch-size-per-gpu>
Please consider citing GC ViT paper if it is useful for your work:
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and Molchanov, Pavlo},
journal={arXiv preprint arXiv:2206.09959},
year={2022}
}
In this section, we list third-party contributions by other users. If you would like to have your work included here, please raise an issue in this repository.
Name | Link | Contributor | Note |
---|---|---|---|
timm | Link | @rwightman | PyTorch |
tfgcvit | Link | @shkarupa-alex | Tensorflow 2.0 (Keras) |
gcvit-tf | Link | @awsaf49 | Tensorflow 2.0 (Keras) |
GCViT-TensorFlow | Link | @EMalagoli92 | Tensorflow 2.0 (Keras) |
Paper Explanation | Link | @awsaf49 | Annotated GC ViT |
Colab Notebook | Link | @awsaf49 | Flower classification |
Kaggle Notebook | Link | @awsaf49 | Flower classification |
Live Demo | Link | @awsaf49 | Hugging Face demo |
-
This repository is built upon the timm library.
-
We would like to sincerely thank the community especially Github users @rwightman, @shkarupa-alex, @awsaf49, @leondgarse, who have provided insightful feedback, which has helped us to further improve GC ViT and achieve even better benchmarks.
Copyright © 2022, NVIDIA Corporation. All rights reserved.
This work is made available under the Nvidia Source Code License-NC. Click here to view a copy of this license.
The pre-trained models are shared under CC-BY-NC-SA-4.0. If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.
For license information regarding the timm repository, please refer to the official website.
For license information regarding the ImageNet dataset, please refer to the official website.