This is an implementation to the CNN model architecture in the paper Accurate Eye Pupil Localization Using Heterogeneous CNN Models
- It was implemented on the public dataset GI4E
- Multiple preprocessing stages were executed on the dataset to reach the required format of predictors and labels for the CNN model
- Finally, the CNN model was trained and tested on ten different images
Reference paper link: https://ieeexplore.ieee.org/abstract/document/8803121
Dataset link: https://www.unavarra.es/gi4e/databases/gi4e/