Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/NVlabs/GCVit

[ICML 2023] Official PyTorch implementation of Global Context Vision Transformers
https://github.com/NVlabs/GCVit

ade20k backbone coco deep-learning imagenet imagenet-classification object-detection pre-train pre-trained-model self-attention semantic-segmentation vision-transformer visual-recognition

Last synced: about 1 month ago
JSON representation

[ICML 2023] Official PyTorch implementation of Global Context Vision Transformers

Host: GitHub
URL: https://github.com/NVlabs/GCVit
Owner: NVlabs
License: other
Created: 2022-06-18T00:44:21.000Z (about 2 years ago)
Default Branch: main
Last Pushed: 2023-12-22T13:04:04.000Z (9 months ago)
Last Synced: 2024-06-19T03:11:29.357Z (3 months ago)
Topics: ade20k, backbone, coco, deep-learning, imagenet, imagenet-classification, object-detection, pre-train, pre-trained-model, self-attention, semantic-segmentation, vision-transformer, visual-recognition
Language: Python
Homepage: https://arxiv.org/abs/2206.09959
Size: 858 KB
Stars: 416
Watchers: 10
Forks: 50
Open Issues: 1
Metadata Files:
- Readme: README.md
- License: LICENSE

Awesome Lists containing this project

README

        # Global Context Vision Transformer (GC ViT)

This repository presents the official PyTorch implementation of **Global Context Vision Transformers** (ICML2023) \

 \

[Global Context Vision

Transformers](https://arxiv.org/pdf/2206.09959.pdf) \

[Ali Hatamizadeh](https://research.nvidia.com/person/ali-hatamizadeh),

[Hongxu (Danny) Yin](https://scholar.princeton.edu/hongxu),

[Greg Heinrich](https://developer.nvidia.com/blog/author/gheinrich/),

[Jan Kautz](https://jankautz.com/), 

and [Pavlo Molchanov](https://www.pmolchanov.com/).

GC ViT  achieves state-of-the-art results across image classification, object detection and semantic segmentation tasks. On ImageNet-1K dataset for classification, GC ViT variants with `51M`, `90M` and `201M` parameters achieve `84.3`, `85.9` and `85.7` Top-1 accuracy, respectively, surpassing comparably-sized prior art such as CNN-based ConvNeXt and ViT-based Swin Transformer.







The architecture of GC ViT is demonstrated in the following:

![gc_vit](https://github.com/NVlabs/GCVit/assets/26806394/86ca853e-56bc-4907-b3e3-0c4611ef9073)

## 💥 News 💥

- **[10.14.2023]** 🔥 We have released the [object detection code](https://github.com/NVlabs/GCVit/tree/main/detection

) !

- **[07.27.2023]**  We will present GC ViT in the (1:30-3:30 HDT) ICML23 session in exhibit hall#1, poster #516.   

- **[07.22.2023]** 🔥🔥 We have released pretrained 21K GC ViT-L checkpoint for 512 x 512 resolution ! 

- **[07.22.2023]** Pretrained checkpoints are now available in official [NVIDIA GCViT HuggingFace](https://huggingface.co/nvidia/GCViT) page !

- **[07.21.2023]** 🔥 We have released the object detection/instance segmentation [code](./detection/README.md) ! 

- **[05.21.2023]** 🔥 We have released ImageNet-21K fine-tuned GC ViT model weights for 224x224 and 384x384.

- **[05.21.2023]** 🔥🔥 We have released new ImageNet-1K GC ViT model weights with **better performance** !

- **[04.24.2023]** 🔥🔥🔥 GC ViT has been accepted to **ICML 2023** !

## Introduction

**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 Benchmarks

**ImageNet-1K Pretrained Models**

  

    Model Variant

    Acc@1

    #Params(M)

    FLOPs(G)

    Download

  

    GC ViT-XXT

    79.9

    12

    2.1

    model

    GC ViT-XT

    82.0

    20

    2.6

    model

    GC ViT-T

    83.5

    28

    4.7

    model

    GC ViT-T2

    83.7

    34

    5.5

    model

    GC ViT-S

    84.3

    51

    8.5

    model

    GC ViT-S2

    84.8

    68

    10.7

    model

    GC ViT-B

    85.0

    90

    14.8

    model

    GC ViT-L

    85.7

    201

    32.6

    model

**ImageNet-21K Pretrained Models**

  

    Model Variant

    Resolution

    Acc@1

    #Params(M)

    FLOPs(G)

    Download

  

    GC ViT-L

    224 x 224

    86.6

    201

    32.6

    model

    GC ViT-L

    384 x 384

    87.4

    201

    120.4

    model

    GC ViT-L

    512 x 512

    87.6

    201

    245.0

    model

## Installation

The dependencies can be installed by running:

```bash

pip install -r requirements.txt

```

## Data Preparation

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:

```bash

  imagenet

  ├── train

  │   ├── class1

  │   │   ├── img1.jpeg

  │   │   ├── img2.jpeg

  │   │   └── ...

  │   ├── class2

  │   │   ├── img3.jpeg

  │   │   └── ...

  │   └── ...

  └── val

      ├── class1

      │   ├── img4.jpeg

      │   ├── img5.jpeg

      │   └── ...

      ├── class2

      │   ├── img6.jpeg

      │   └── ...

      └── ...

 

  ```

## Commands

### Training on ImageNet-1K From Scratch (Multi-GPU)

The `GC ViT` model can be trained on ImageNet-1K dataset by running:

```bash

python -m torch.distributed.launch --nproc_per_node  --master_port 11223  train.py \ 

--config  --data_dir  --batch-size --amp  --tag  --model-ema

```

To resume training from a pre-trained checkpoint:

```bash

python -m torch.distributed.launch --nproc_per_node  --master_port 11223  train.py \ 

--resume  --config  --amp --data_dir  --batch-size  --tag  --model-ema

```

### Evaluation

To evaluate a pre-trained checkpoint using ImageNet-1K validation set on a single GPU:

```bash

python validate.py --model  --checkpoint  --data_dir  --batch-size 

```

## Citation

Please consider citing GC ViT paper if it is useful for your work:

```

@inproceedings{hatamizadeh2023global,

  title={Global context vision transformers},

  author={Hatamizadeh, Ali and Yin, Hongxu and Heinrich, Greg and Kautz, Jan and Molchanov, Pavlo},

  booktitle={International Conference on Machine Learning},

  pages={12633--12646},

  year={2023},

  organization={PMLR}

}

```

## Third-party Implementations and Resources

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 | Framework

|:---:|:---:|:---:|:---------:|

|timm|[Link](https://github.com/rwightman/pytorch-image-models)| @rwightman | PyTorch

|tfgcvit|[Link](https://github.com/shkarupa-alex/tfgcvit)| @shkarupa-alex | Tensorflow 2.0 (Keras)

|gcvit-tf|[Link](https://github.com/awsaf49/gcvit-tf)| @awsaf49 | Tensorflow 2.0 (Keras)

|GCViT-TensorFlow|[Link](https://github.com/EMalagoli92/GCViT-TensorFlow)| @EMalagoli92 | Tensorflow 2.0 (Keras)

|keras_cv_attention_models|[Link](https://github.com/leondgarse/keras_cv_attention_models/tree/main/keras_cv_attention_models/gcvit)| @leondgarse | Keras

|flaim|[Link](https://github.com/BobMcDear/flaim)| @BobMcDear | JAX/Flax

## Additional Resources

We list additional GC ViT resources such as notebooks, demos, paper explanations in this section. If you have created similar items and would like to be included, please raise an issue in this repository.

| Name | Link | Contributor | Note

|:---:|:---:|:---:|:---------:|

|Paper Explanation|[Link](https://www.kaggle.com/code/awsaf49/guie-global-context-vit-gcvit)| @awsaf49 | Annotated GC ViT

|Colab Notebook|[Link](https://colab.research.google.com/github/awsaf49/gcvit-tf/blob/main/notebooks/GCViT_Flower_Classification.ipynb)| @awsaf49 | Flower classification

|Kaggle Notebook|[Link](https://www.kaggle.com/code/awsaf49/flower-classification-gcvit-global-context-vit/notebook)| @awsaf49 | Flower classification

|Live Demo|[Link](https://huggingface.co/spaces/awsaf49/gcvit-tf)| @awsaf49 | Hugging Face demo

## Licenses

Copyright © 2023, NVIDIA Corporation. All rights reserved.

This work is made available under the Nvidia Source Code License-NC. Click [here](LICENSE) to view a copy of this license.

The pre-trained models are shared under [CC-BY-NC-SA-4.0](https://creativecommons.org/licenses/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, please refer to its [repository](https://github.com/rwightman/pytorch-image-models).

For license information regarding the ImageNet dataset, please refer to the ImageNet [official website](https://www.image-net.org/). 

## Acknowledgement

- This repository is built upon the [timm](https://github.com/rwightman/pytorch-image-models) 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.