This repository is home to my exploration of lightweight CNN architectures tailored for image classification tasks. It serves as a companion to the research paper, providing practical insights into CNN optimizations for models with restricted parameter budgets.
Included in this repo are several models:
- Base Model: The foundational model of my investigation.
- Models V1-V6: These are my experimental models, each incorporating unique architectural features.
- Model Alpha: The pinnacle of the experimentation, showcasing the combined learnings from previous models.
I encourage you to fork the repo, suggest enhancements, or utilize our models for your research. Collaboration is key to advancing the field.
If you have questions or would like to discuss the research further, please reach out.