Focal_Transformer

Focal Self-attention for Local-Global Interactions in Vision Transformers, arxiv

PaddlePaddle training/validation code and pretrained models for Focal Transformer.

The official pytorch implementation is here.

This implementation is developed by PaddleViT.

Focal Transformer Model Overview

Update

• Update(2022-04-11): Code is updated.
• Update(2021-10-21): Code is released and ported weights are uploaded.

Models Zoo

Model	Acc@1	Acc@5	#Params	FLOPs	Image Size	Crop_pct	Interpolation	Link
Focal-T	82.03	95.86	28.9M	4.9G	224	0.875	bicubic	google/baidu
Focal-T (use conv)	82.70	96.14	30.8M	4.9G	224	0.875	bicubic	google/baidu
Focal-S	83.55	96.29	51.1M	9.4G	224	0.875	bicubic	google/baidu
Focal-S (use conv)	83.85	96.47	53.1M	9.4G	224	0.875	bicubic	google/baidu
Focal-B	83.98	96.48	89.8M	16.4G	224	0.875	bicubic	google/baidu
Focal-B (use conv)	84.18	96.61	93.3M	16.4G	224	0.875	bicubic	google/baidu

*The results are evaluated on ImageNet2012 validation set.

Data Preparation

ImageNet2012 dataset is used in the following file structure:

│imagenet/
├──train_list.txt
├──val_list.txt
├──train/
│  ├── n01440764
│  │   ├── n01440764_10026.JPEG
│  │   ├── n01440764_10027.JPEG
│  │   ├── ......
│  ├── ......
├──val/
│  ├── n01440764
│  │   ├── ILSVRC2012_val_00000293.JPEG
│  │   ├── ILSVRC2012_val_00002138.JPEG
│  │   ├── ......
│  ├── ......

• train_list.txt: list of relative paths and labels of training images. You can download it from: google/baidu
• val_list.txt: list of relative paths and labels of validation images. You can download it from: google/baidu

Usage

To use the model with pretrained weights, download the .pdparam weight file and change related file paths in the following python scripts. The model config files are located in ./configs/.

For example, assume weight file is downloaded in ./focal_base_patch4_window7_224.pdparams, to use the focal_base_patch4_window7_224 model in python:

from config import get_config
from focal_transformer import build_focal as build_model
# config files in ./configs/
config = get_config('./configs/focal_base_patch4_window7_224.yaml')
# build model
model = build_model(config)
# load pretrained weights
model_state_dict = paddle.load('./focal_base_patch4_window7_224.pdparams')
model.set_state_dict(model_state_dict)

Evaluation

To evaluate model performance on ImageNet2012, run the following script using command line:

sh run_eval_multi.sh

or

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
python main_multi_gpu.py \
-cfg='./configs/focal_base_patch4_window7_224.yaml' \
-dataset='imagenet2012' \
-batch_size=256 \
-data_path='/dataset/imagenet' \
-eval \
-pretrained='./focal_base_patch4_window7_224.pdparams' \
-amp

Note: if you have only 1 GPU, change device number to CUDA_VISIBLE_DEVICES=0 would run the evaluation on single GPU.

Training

To train the model on ImageNet2012, run the following script using command line:

sh run_train_multi.sh

or

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
python main_multi_gpu.py \
-cfg='./configs/focal_base_patch4_window7_224.yaml' \
-dataset='imagenet2012' \
-batch_size=256 \
-data_path='/dataset/imagenet' \
-amp

Note: it is highly recommanded to run the training using multiple GPUs / multi-node GPUs.

Arguments

• -cfg: path of model config file (.yaml), stored in ./configs.
• -dataset: dataset name, e.g., imagenet2012, cifar10, cifar100.
• -data_path: dataset folder path
• -batch_size: batch size，default: 32.
• -image_size: input image size，default224.
• -num_classes: number of classes, default: 1000.
• -output: output folder for storing weights and logs，default: ./output.
• -pretrained: pretrain model weights file path, (.pdparams file ext is NOT needed) default: None.
• -resume: resume model weight and opt file path, (.paparams and .pdopts file ext are NOT needed, default: None.
• -last_epoch: start epoch，default: None.
• -save_freq: number of epochs to save checkpoint，default: 1.
• -log_freq: number of iters to print logging，default: 100.
• -validate_freq: number of epochs to do validation during training，default: 10.
• -accum_iter: number of iteration for iter accumulation, default: 1.
• -num_workers: number of workers for data loading，default: 1.
• -ngpus: number of GPUs to use，you can control GPUs by CUDA_VISIBLE_DEVICES, just set this to -1 default: -1.
• -eval: start eval mode.
• -amp: start amp training.

-cfg,-dataset and -data_path in main_single_gpu.py and main_multi_gpu.py are MUST-HAVE settings.

Reference

@misc{yang2021focal,
    title={Focal Self-attention for Local-Global Interactions in Vision Transformers}, 
    author={Jianwei Yang and Chunyuan Li and Pengchuan Zhang and Xiyang Dai and Bin Xiao and Lu Yuan and Jianfeng Gao},
    year={2021},
    eprint={2107.00641},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}