Implementation of YOLOv2 for real time object detection.
- anchors.txt contains the dimentions of the anchor boxes.
- data.csv contains metadata having 3 columns: ImageLoc , ROIs, Classes. Each row of Data.csv contains: path_to_the_image_on_disk, [xmin, ymin, height, width], classes
Steps for running the model:
- Open terminal
- pip install requirements.txt --> for python 2.x |or| pip3 install requirements.txt --> for python 3.x
- python maybe_download_and_extract.py
- python make_tfrecord.py
- Run jupyter notebook using command: jupyter-notebook and launch model.ipynb