Animal Recognition

Code developed in AI class as an exercise with image classification.

This work compares KNN method with ResNet. The results shows that KNN method got higher accuracy than ResNet.

Method	Accuracy
KNN k=1	97.69%
KNN k=3	98.46%
KNN k=5	98.46%
KNN k=7	98.46%
KNN k=11	98.46%
ResNet	91%

Python Libraries

• sentence-transformers (pip install sentence-transformers)

• imbalanced-learn (pip install -U imbalanced-learn)

• ktrain (pip install ktrain)

Members

Gabriela Barion Vidal