Skip to content

Latest commit

 

History

History
152 lines (114 loc) · 6.77 KB

README.md

File metadata and controls

152 lines (114 loc) · 6.77 KB
Raw

HEtero-Assists Distillation (ECCV 2022)

    _/    _/  _/_/_/_/    _/_/    _/_/_/
   _/    _/  _/        _/    _/  _/    _/
  _/_/_/_/  _/_/_/    _/_/_/_/  _/    _/
 _/    _/  _/        _/    _/  _/    _/
_/    _/  _/_/_/_/  _/    _/  _/_/_/

This repository is the official implementation of "HEtero-Assists Distillation for Heterogeneous Object Detectors".

lint wakatime

Preparation

Download the MS-COCO dataset to data/coco.

Download MMDetection pretrained models to pretrained/mmdetection

mkdir -p pretrained/mmdetection
wget -P pretrained/mmdetection https://download.openmmlab.com/mmdetection/v2.0/faster_rcnn/faster_rcnn_r50_fpn_mstrain_3x_coco/faster_rcnn_r50_fpn_mstrain_3x_coco_20210524_110822-e10bd31c.pth
wget -P pretrained/mmdetection https://download.openmmlab.com/mmdetection/v2.0/third_party/mobilenet_v2_batch256_imagenet-ff34753d.pth
wget -P pretrained/mmdetection https://download.openmmlab.com/mmdetection/v2.0/retinanet/retinanet_r50_fpn_mstrain_3x_coco/retinanet_r50_fpn_mstrain_3x_coco_20210718_220633-88476508.pth

Download torchvision pretrained models to pretrained/torchvision

mkdir -p ~/.cache/torch/hub/checkpoints
ln -s ~/.cache/torch/hub/checkpoints pretrained/torchvision
wget -P pretrained/torchvision https://download.pytorch.org/models/resnet18-f37072fd.pth
wget -P pretrained/torchvision https://download.pytorch.org/models/resnet50-0676ba61.pth

The directory tree should be like this

HEAD
├── data
│   └── coco -> ~/Developer/datasets/coco
│       ├── annotations
│       │   ├── instances_train2017.json
│       │   └── instances_val2017.json
│       ├── train2017
│       │   └── ...
│       └── val2017
│           └── ...
├── pretrained
│   ├── mmdetection
│   │   ├── faster_rcnn_r50_fpn_mstrain_3x_coco_20210524_110822-e10bd31c.pth
│   │   ├── mobilenet_v2_batch256_imagenet-ff34753d.pth
│   │   └── retinanet_r50_fpn_mstrain_3x_coco_20210718_220633-88476508.pth
│   └── torchvision -> ~/.cache/torch/hub/checkpoints
│       ├── resnet18-f37072fd.pth
│       └── resnet50-0676ba61.pth
└── ...

Installation

Create a conda environment and activate it.

conda create -n HEAD python=3.8
conda activate HEAD

Install MMDetection following the official instructions. For example,

pip install torch==1.9.1+cu102 torchvision==0.10.1+cu102 -f https://download.pytorch.org/whl/torch_stable.html
pip install -U openmim
mim install mmcv_full==1.4.6
pip install mmdet==2.20

Install todd.

pip install todd_ai==0.2.3a9 -i https://pypi.org/simple

Note that the requirements.txt is not intended for users. Please follow the above instructions.

Training

python tools/train.py configs/head/head_retina_faster_r18_fpn_mstrain_1x_coco.py --work-dir work_dirs/debug --seed 3407

For distributed training

bash tools/dist_train.sh configs/head/head_retina_faster_r18_fpn_mstrain_1x_coco.py 8 --work-dir work_dirs/debug --seed 3407

Results

All logs and checkpoints can be found in the Google Drive.

Baseline

Method Student Teacher mAP Config Comment
FitNet R18 RetinaNet R50 RetinaNet $35.6$ fitnet_retina with weight transfer

HEAD

Teachers and students are all trained with multi-scale, for 3x and 1x schedulers respectively.

Student Teacher Assist AKD CKD mAP Config
R18 RetinaNet $31.7$ refer to mmdetection
R18 RetinaNet R50 Faster R-CNN $\checkmark$ $33.4$ retina_faster_r18
R18 RetinaNet R50 Faster R-CNN $\checkmark$ $\checkmark$ $35.7$ HEAD_dag_retina_faster_r18
R18 RetinaNet R50 Faster R-CNN $\checkmark$ $\checkmark$ $\checkmark$ $36.1$ HEAD_retina_faster_r18
MNv2 RetinaNet $27.8$ retinanet_mnv2
MNv2 RetinaNet R50 Faster R-CNN $\checkmark$ $28.9$ retina_faster_mnv2
MNv2 RetinaNet R50 Faster R-CNN $\checkmark$ $\checkmark$ $32.2$ HEAD_dag_retina_faster_mnv2
MNv2 RetinaNet R50 Faster R-CNN $\checkmark$ $\checkmark$ $\checkmark$ $33.1$ HEAD_retina_faster_mnv2

TF-HEAD

Coming soon...

Developer Guides

Setup

pip install https://download.pytorch.org/whl/cpu/torch-1.9.1-cp38-none-macosx_11_0_arm64.whl
pip install https://download.pytorch.org/whl/cpu/torchvision-0.10.0-cp38-cp38-macosx_11_0_arm64.whl
pip install -e ./../mmcv
pip install mmdet==2.20
pip install commitizen
pip install -U pre-commit
pre-commit install
pre-commit install -t commit-msg

TODO

  • complete distributed train/test guide
  • more configs
  • etc.