Date of Publication :15th December 2024

Abstract: The dependability of AI and machine learning models in fast-evolving fields relies on having access to thorough and high-caliber data. Nevertheless, the problem of missing or incomplete data is widespread and can greatly reduce the accuracy and effectiveness of trained models. This study is centered on tackling this difficulty by using MISGAN, a new method created to produce missing images in datasets. The goal is to create lifelike and top-notch images that smoothly fill in the areas left empty due to missing data. This approach enhances the value of existing datasets and contributes to the development of more accurate an d reliable machine learning models by leveraging the power of MISGANs.

Reference :