The dataset is from Kaggle's Public Dataset repository. It contains images of natural scences around the world. The goal is to classify surrounding scences which includes categories like buildings, forest, glacier, mountain, sea , street.
There are around 14k images in Train, 3k in Test and 7k in Prediction. Images are of size 150x150 with RGB Channel. For each class, there are around ~2k images.
Used Keras's Image Generator Method to Augument Images during Training.
| Model | Optimizer | Epochs | Validation Accuracy | Train Accuracy |
|---|---|---|---|---|
| LeNet-5 | RMSProp | 100 | 0.65 | 0.58 |
| AlexNet | Adam | 100 | 0.70 | 0.69 |
| VGG-16 | SGD | 100 | 0.63 | 0.66 |