The increase in digitally altered images creates a big problem for regular face detection systems, as altered faces can easily avoid detection. This study introduces a new method to improve face detection accuracy with altered images using advanced deep learning techniques. Our model uses a selected dataset of altered faces for training and validation. The deep neural network includes YoloV8, CNN, VGG16, RESNET50, and Inceptionv3 to accurately differentiate between real and altered facial features. Test results show the effectiveness of our method, achieving high accuracy. We also compare our approach with current face detection methods.