This is a sample code about yolo-lite.
pip install -r requirement.txt
default parameter use "python train -h"
python train.py --mode train --pretrain_model=./pretrained/cp-30-3.614092 --se --bn
python evaluate.py --pretrain_model=./pretrained/cp-30-3.614092 --se --bn # hack api depents on coco-tools
python demo.py --mode video --pretrain_model=./pretrained/cp-30-3.614092 --se --bn python demo.py --mode video --pretrain_model=./pretrained/cp-145-4.073046
1. coco-pretrain model
2. some serilization-bug between keras and tf2.1
cant load model with partial. (as far known)
3. other bug
- https://arxiv.org/pdf/1811.05588
- https://towardsdatascience.com/review-senet-squeeze-and-excitation-network-winner-of-ilsvrc-2017-image-classification-a887b98b2883
- https://github.com/YunYang1994/tensorflow-yolov3
- https://github.com/cocodataset/cocoapi/blob/e140a084d678eacd18e85a9d8cfa45d1d5911db9/PythonAPI/pycocotools/coco.py