Created a dataset to detect Excited and Bored Features which using YOLOV5 model.

Setup Pytorch library compatible with your system requirements https://pytorch.org/get-started/locally/

Setup the YOLOV5 directory https://github.com/ultralytics/yolov5

LabelImg to create Annotation boxes https://github.com/tzutalin/labelImg

1. Git Clone the YOLOV5 and llabelImg Repositories
2. Create a data/images and data/labels folder in the workspace
3. Run the script. The current script was built to be inline with on Jupyter Notebook in Anaconda Navigator on Windows 11.

Trined the models through 500 epochs in batches of 16 to give optimized model with 85 percent accuracy.

Bored Detection

Excited Detection