Repo containing different Deep Learning Projects in Image Classification that I have made.
In this notebook, I bulit a deep learning computer vision model to identify over 100 different types of food, with a total of 101,000 images (75,750 training images and 25,250 testing images).
Target Score:
- The Goal is to beat the DeepFood paper (77.4% top-1 accuracy).
- Alongside the Food101 paper (50.76% top-1 accuracy).
That were trained on the same dataset.
Model Score:
- validation_accuracy on previously unseen data of 79.92%.
Passed Target Score: True
In this notebook, I made a computer vision neural networks to build an image classification model capable of predicting over 100 different dog breeds.
Model Score:
- Multi Class Log Loss on Kaggle's test dataset of 4.78.
In this notebook, I bulit a deep learning computer vision model to classifiy fruits images into 6 categories:
- Fresh Apples
- Fresh Bananas
- Fresh Oranges
- Rotten Apples
- Rotten Bananas
- Rotten Oranges
Model Score:
- validation_accuracy on previously unseen data of 98.45%.