Apple or Tomato? Binary Image Classification [TENSORFLOW]
In this project, we will make a binary classification model that can identify whether it is an apple or a tomato in the image. Apples and tomatoes share some common characteristics such as the round shape or the red color. It might be a tricky work to train a model properly, so that it can differntiate the two different obejcts.
....
This is a basic exercise to develop a image classifier with Tensorflow. It can be used as a basic guidline to create a model from a pretrained one.
The dataset can be downloaded from https://www.kaggle.com/datasets/samuelcortinhas/apples-or-tomatoes-image-classification/code?datasetId=2702329&sortBy=voteCount
To summarize the result, we were able to develop a model that has made a quite good prediction with the test dataset. The accuracy score was 92%. We first experimented with our fine-tuned pretrained model of InceptionResnet, different pretrained models can be used to make further experiments
correction => input shape is 3 dimensional, meaning -> 299x299x3. 3 is because the image is a RGB color image