Face Detection and Alignment using Multi-task Cascaded Convolutional Networks (MTCNN)

MTCNN-TensorFlow - Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks based Counting

Face Detection and Alignment-68 using MTCNN

A Keras implementation of MTCNN from [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://kpzhang93.github.io/MTCNN_face_detection_alignment/paper/spl.pdf)

No description available

MTCNN-TensorFlow - Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks

Using MTCNN and various pre-processing techniques to optimise Face Detection and Alignment for Face Recognition.

This work is used for reproduce MTCNN,a Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks.

Face detection and one-to-one verification using MTCNN for bounding boxes/keypoints and FaceNet embeddings with cosine similarity, plus alignment utilities, threshold tuning, and evaluation (confusion matrix, ROC); research/education only.

This repository implements a face grouping system that automatically clusters photos of the same person, similar to how Google Photos works. It uses a hybrid embedding approach by combining multiple pretrained models ArcFace, FaceNet, SFace, VGG-Face with MTCNN alignment for consistent face detection.

The project focuses on detecting deepfake images using transfer learning with MobileNetV2 fine-tuned on a dataset of real and fake images. MTCNN is used for face detection and alignment. Data augmentation techniques are applied to increase dataset diversity. The deepfake detection model consists of MobileNetV2,Dense layers for binary classification