/KNNDataCleaner

KNNDataCleaner: Streamlining Image Datasets for Enhanced Machine Learning Performance.

Primary LanguageC++MIT LicenseMIT

KNNDataCleaner

KNNDataCleaner: Remove duplicate images from the dataset using the KNN algorithm, which may cause data imbalance

For the successful operation of the KNNDataCleaner, it is essential to organize your image dataset into specifically named folders. Each folder should be named numerically in a sequential order, such as 1, 2, 3, 4, 5, 6, 7, etc. This naming convention is crucial for the program to accurately read and process the images from the dataset. Ensure that the images belonging to a particular category are placed in the corresponding numbered folder to facilitate smooth and error-free processing