This dataset contains total of 21,122 fully labeled images, featuring 20 different kinds of fruits. It is structured into an 80% training set (16,899 images) and a 20% testing set (4,223 images), facilitating a ready-to-use framework for model training and evaluation.
Additionally, there are two CSV files that label the types of fruits depicted in each image.
