Rohan Vij, Kevin Geng, Shan Kanwar - Dublin High School
Josh Tittiranonda - Stanford Online High School
You can view the Jupyter Notebook for a tutorial on training and inferencing here.
Documentation (paper, video, spreadsheet, etc) can be found here.
The trained models (and their associated testing videos) can be found here.
Note that we had severe limitations with the final capability of the model - we had to use Google Colab due to a lack of home/school training resources and were not able to spend much money to train our models for very long.
YOLOv7's inbuilt calculations recommended a minimum of 171 epochs to train the model on all of the data we provided it - the most we were able to train was 100 epochs (58.4% of the minimum required for decent performance).
When training the model yourself, try to run up to 171 epochs if your compute time and budget allow.
This project would not have been possible without the support of the open-source community and the creators of YOLOv7.
This project is licensed under the MIT License. See the LICENSE file for details.