I acknowledged that I co-coperated with @Camille Ko. We had the same contribution to this project, including code and report.
The task is to classify animals in the animal dataset which have has 6,270 images categorized into 151 different labeled classes.
We used the Animal dataset. All images are RGB images of 224 pixels wide by 224 pixels high in .jpg format. The images are separated in 151 folders according to their respective class.
| Version | Accuracy | Efficiency |
|---|---|---|
| best model | 98.28%🚀 | 0.43G〽️ |
| base model | 37.53% | 0.69G |
To understand how to achieve this perfromance, please refer to report.