BorderDet

This project provides an implementation for "BorderDet: Border Feature for Dense Object Detection" (ECCV2020 Oral) on PyTorch.

For the reason that experiments in the paper were conducted using internal framework, this project reimplements them on cvpods and reports detailed comparisons below.

Requirements

• cvpods

Get Started

• install cvpods locally (requires cuda to compile)

python3 -m pip install 'git+https://github.com/Megvii-BaseDetection/cvpods.git'
# (add --user if you don't have permission)

# Or, to install it from a local clone:
git clone https://github.com/Megvii-BaseDetection/cvpods.git
python3 -m pip install -e cvpods

# Or,
pip install -r requirements.txt
python3 setup.py build develop

• prepare datasets

cd /path/to/cvpods
cd datasets
ln -s /path/to/your/coco/dataset coco

• Train & Test

git clone https://github.com/Megvii-BaseDetection/BorderDet.git
cd BorderDet/playground/detection/coco/borderdet/borderdet.res50.fpn.coco.800size.1x  # for example

Train

pods_train --num-gpus 8

Test

pods_test --num-gpus 8 \
    MODEL.WEIGHTS /path/to/your/save_dir/ckpt.pth # optional
    OUTPUT_DIR /path/to/your/save_dir # optional

Multi node training

sudo apt install net-tools ifconfig

pods_train --num-gpus 8 --num-machines N --machine-rank 0/1/.../N-1 --dist-url "tcp://MASTER_IP:port"

Results on COCO

For your convenience, we provide the performance of the following trained models. All models are trained with 16 images in a mini-batch and frozen batch normalization. All model including X_101/DCN_X_101 will be released soon.

Model	Multi-scale training	Multi-scale testing	Testing time / im	AP (minival)	Link
FCOS_R_50_FPN_1x	No	No	54ms	38.7	download
BD_R_50_FPN_1x	No	No	60ms	41.4	download
BD_R_101_FPN_1x	Yes	No	76ms	45.0	download
BD_X_101_32x8d_FPN_1x	Yes	No	124ms	45.6	download
BD_X_101_64x4d_FPN_1x	Yes	No	123ms	46.2	download
BD_DCNV2_X_101_32x8d_FPN_1x	Yes	No	150ms	47.9	download
BD_DCNV2_X_101_64x4d_FPN_1x	Yes	No	156ms	47.5	download

Acknowledgement

cvpods is developed based on Detectron2. For more details about official detectron2, please check DETECTRON2.

Contributing to the project

Any pull requests or issues are welcome.