👀
One Sequence
You Only -
TL;DR: We study the transferability of the vanilla ViT pre-trained on mid-sized ImageNet-1k to the more challenging COCO object detection benchmark.
-
This project is under active development.
You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection
by Yuxin Fang1 *, Bencheng Liao1 *, Xinggang Wang1
✉️ , Jiemin Fang2, 1, Jiyang Qi1, Rui Wu3, Jianwei Niu3, Wenyu Liu1.1 School of EIC, HUST, 2 Institute of AI, HUST, 3 Horizon Robotics.
(*) equal contribution, (
✉️ ) corresponding author.arXiv technical report (arXiv 2106.00666)
You Only Look at One Sequence (YOLOS)
The Illustration of YOLOS
Highlights
Directly inherited from ViT (DeiT), YOLOS is not designed to be yet another high-performance object detector, but to unveil the versatility and transferability of Transformer from image recognition to object detection. Concretely, our main contributions are summarized as follows:
-
We use the mid-sized
ImageNet-1k
as the sole pre-training dataset, and show that a vanilla ViT (DeiT) can be successfully transferred to perform the challenging object detection task and produce competitiveCOCO
results with the fewest possible modifications, i.e., by only looking at one sequence (YOLOS). -
We demonstrate that 2D object detection can be accomplished in a pure sequence-to-sequence manner by taking a sequence of fixed-sized non-overlapping image patches as input. Among existing object detectors, YOLOS utilizes minimal 2D inductive biases. Moreover, it is feasible for YOLOS to perform object detection in any dimensional space unaware the exact spatial structure or geometry.
-
For ViT (DeiT), we find the object detection results are quite sensitive to the pre-train scheme and the detection performance is far from saturating. Therefore the proposed YOLOS can be used as a challenging benchmark task to evaluate different pre-training strategies for ViT (DeiT).
-
We also discuss the impacts as wel as the limitations of prevalent pre-train schemes and model scaling strategies for Transformer in vision through transferring to object detection.
Results
Model | Pre-train Epochs | ViT (DeiT) Weight / Log | Fine-tune Epochs | Eval Size | YOLOS Checkpoint / Log | AP @ COCO val |
---|---|---|---|---|---|---|
YOLOS-Ti |
300 | FB | 300 | 512 | Baidu Drive, Google Drive / Log | 28.7 |
YOLOS-S |
200 | Baidu Drive, Google Drive / Log | 150 | 800 | Baidu Drive, Google Drive / Log | 36.1 |
YOLOS-S |
300 | FB | 150 | 800 | Baidu Drive, Google Drive / Log | 36.1 |
YOLOS-S (dWr) |
300 | Baidu Drive, Google Drive / Log | 150 | 800 | Baidu Drive, Google Drive / Log | 37.6 |
YOLOS-B |
1000 | FB | 150 | 800 | Baidu Drive, Google Drive / Log | 42.0 |
Notes:
- The access code for
Baidu Drive
isyolo
. - The
FB
stands for model weights provided by DeiT (paper, code). Thanks for their wonderful works. - We will update other models in the future, please stay tuned :)
Requirement
This codebase has been developed with python version 3.6, PyTorch 1.5+ and torchvision 0.6+:
conda install -c pytorch pytorch torchvision
Install pycocotools (for evaluation on COCO) and scipy (for training):
conda install cython scipy
pip install -U 'git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI'
Data preparation
Download and extract COCO 2017 train and val images with annotations from http://cocodataset.org. We expect the directory structure to be the following:
path/to/coco/
annotations/ # annotation json files
train2017/ # train images
val2017/ # val images
Training
Before finetuning on COCO, you need download the ImageNet pretrained model to the /path/to/YOLOS/
directory
To train the YOLOS-Ti
model in the paper, run this command:
python -m torch.distributed.launch \
--nproc_per_node=8 \
--use_env main.py \
--coco_path /path/to/coco
--batch_size 2 \
--lr 5e-5 \
--epochs 300 \
--backbone_name tiny \
--pre_trained /path/to/deit-tiny.pth\
--eval_size 512 \
--init_pe_size 800 1333 \
--output_dir /output/path/box_model
To train the YOLOS-S
model with 200 epoch pretrained Deit-S in the paper, run this command:
python -m torch.distributed.launch
--nproc_per_node=8
--use_env main.py
--coco_path /path/to/coco --batch_size 1
--lr 2.5e-5
--epochs 150
--backbone_name small
--pre_trained /path/to/deit-small-200epoch.pth
--eval_size 800
--init_pe_size 512 864
--mid_pe_size 512 864
--output_dir /output/path/box_model
To train the YOLOS-S
model with 300 epoch pretrained Deit-S in the paper, run this command:
python -m torch.distributed.launch \ --nproc_per_node=8 \ --use_env main.py \ --coco_path /path/to/coco --batch_size 1 \ --lr 2.5e-5 \ --epochs 150 \ --backbone_name small \ --pre_trained /path/to/deit-small-300epoch.pth\ --eval_size 800 \ --init_pe_size 512 864 \ --mid_pe_size 512 864 \ --output_dir /output/path/box_model
To train the YOLOS-S (dWr)
model in the paper, run this command:
python -m torch.distributed.launch \
--nproc_per_node=8 \
--use_env main.py \
--coco_path /path/to/coco
--batch_size 1 \
--lr 2.5e-5 \
--epochs 150 \
--backbone_name small_dWr \
--pre_trained /path/to/deit-small-dWr-scale.pth\
--eval_size 800 \
--init_pe_size 512 864 \
--mid_pe_size 512 864 \
--output_dir /output/path/box_model
To train the YOLOS-B
model in the paper, run this command:
python -m torch.distributed.launch \
--nproc_per_node=8 \
--use_env main.py \
--coco_path /path/to/coco
--batch_size 1 \
--lr 2.5e-5 \
--epochs 150 \
--backbone_name base \
--pre_trained /path/to/deit-base.pth\
--eval_size 800 \
--init_pe_size 800 1344 \
--mid_pe_size 800 1344 \
--output_dir /output/path/box_model
Evaluation
To evaluate YOLOS-Ti
model on COCO, run:
python main.py --coco_path /path/to/coco --batch_size 2 --backbone_name tiny --eval --eval_size 512 --init_pe_size 800 1333 --resume /path/to/YOLOS-Ti
To evaluate YOLOS-S
model on COCO, run:
python main.py --coco_path /path/to/coco --batch_size 1 --backbone_name small --eval --eval_size 800 --init_pe_size 512 864 --mid_pe_size 512 864 --resume /path/to/YOLOS-S
To evaluate YOLOS-S (dWr)
model on COCO, run:
python main.py --coco_path /path/to/coco --batch_size 1 --backbone_name small_dWr --eval --eval_size 800 --init_pe_size 512 864 --mid_pe_size 512 864 --resume /path/to/YOLOS-S(dWr)
To evaluate YOLOS-B
model on COCO, run:
python main.py --coco_path /path/to/coco --batch_size 1 --backbone_name small --eval --eval_size 800 --init_pe_size 800 1344 --mid_pe_size 800 1344 --resume /path/to/YOLOS-B
Visualization
We have observed some intriguing properties of YOLOS, and we are working on a notebook to better demonstrate them, please stay tuned :)
Visualize box prediction and object categories distribution:
- To Get visualization in the paper, you need the finetuned YOLOS models on COCO, run following command to get 100 Det-Toks prediction on COCO val split, then it will generate
/path/to/YOLOS/visualization/modelname-eval-800-eval-pred.json
python cocoval_predjson_generation.py --coco_path /path/to/coco --batch_size 1 --backbone_name small --eval --eval_size 800 --init_pe_size 512 864 --mid_pe_size 512 864 --resume /path/to/yolos-s-model.pth --output_dir ./visualization
- To get all ground truth object categories on all images from COCO val split, run following command to generate
/path/to/YOLOS/visualization/coco-valsplit-cls-dist.json
python cocoval_gtclsjson_generation.py --coco_path /path/to/coco --batch_size 1 --output_dir ./visualization
- To visualize the distribution of Det-Toks' bboxs and categories, run following command to generate
.png
files in/path/to/YOLOS/visualization/
python visualize_dettoken_dist.py --visjson /path/to/YOLOS/visualization/modelname-eval-800-eval-pred.json --cococlsjson /path/to/YOLOS/visualization/coco-valsplit-cls-dist.json
Visualize self-attention of the [DetTok] token on the different heads of the last layer:
we are working on a notebook to better demonstrate them, please stay tuned :)
❤️
Acknowledgement This project is based on DETR (paper, code), DeiT (paper, code) and timm. Thanks for their wonderful works.
Citation
If you find our paper and code useful in your research, please consider giving a star
@article{YOLOS,
title={You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection},
author={Fang, Yuxin and Liao, Bencheng and Wang, Xinggang and Fang, Jiemin and Qi, Jiyang and Wu, Rui and Niu, Jianwei and Liu, Wenyu},
journal={arXiv preprint arXiv:2106.00666},
year={2021}
}