Thundernet-pytorch

thundernet object detection

dataset

we used COCO2017 dataset to training, dataset structure:

COCO
├── annotations
│   ├── instances_train2017.json
│   └── instances_val2017.json
│── images
    ├── train2017
    └── val2017

running script:

bash train.sh

Name		Name	Last commit message	Last commit date
Latest commit History 61 Commits
data		data
src		src
thundernet		thundernet
utils		utils
README.md		README.md
detector.py		detector.py
load_data.py		load_data.py
requirements.txt		requirements.txt
train.py		train.py
train.sh		train.sh