Faster R-CNN R-101-FPN model was implemented with tensorflow 2.0.
- python 3.*
- tensorflow 2.*
- cv2
- ...
-
Step1. Download pretrained model from Google Drive or 百度网盘 and put it in
weights/faster_rcnn.h5
directory. -
Step2.
pip install required packages
according to your system prompts. -
Step3.
python train_model.py
- Step4. You may need to check
train_model.ipynb
andinspect_model.ipynb
for some insight Since Faster-RCNN is a really huge project.
This work builds on many excellent works, which include:
- Heavily based on tf-eager-fasterrcnn
- matterport/Mask_RCNN
- open-mmlab/mmdetection