[Python] Why is my face recognition model detecting downloaded and live camera images but not...

Small Info About the Project: This project is a Face Recognition System that can analyze images from both saved image files and webcam input. The model detects whether a given image is from a live camera, a downloaded image, or a screenshot. It uses face embeddings from InsightFace, along with image analysis techniques such as edge detection, brightness measurement, and frequency domain analysis (FFT) to differentiate between these image types. The system also classifies faces as known or unknown based on a trained dataset. But when I give a path to a screenshot image stored in my folder, this model not predicting as screenshot image, instead it is detecting it as a live camera image. How to resolve this error detection?

Continue reading...