Yolov3

Dataset Used:

Pedestrian Detection Data set

https://www.kaggle.com/datasets/karthika95/pedestrian-detection/download?datasetVersionNumber=1

Missing files:

yolov3.weights

https://pjreddie.com/media/files/yolov3.weights

Dependencies:

pip3 install -r requirements.txt

Steps to run:

Download dataset from link
Download yolov3.weights from the link
Pip install dependecies
Create python3 virtual environment and run-  pip3 install -r requirements.txt


iou.py -> to calculate iou for predicted boxes and ground truth

Run these sequentially
python3 yolov3_to_csv.py             This creates an output.csv of model predictions, save it in a different path
python3 yolov3_to_csv_full.py        This opens ouput.csv and does mapping of persons in image predictions to ground truth
python3 iou_seperate.py              To view the bounding boxes around each individual person in the image

python3 iou_sampleBox.py             To view the iou mapping of each image

Change image path wherever necessary to run locally from downloaded dataset