Skip to content

Latest commit

 

History

History
15 lines (11 loc) · 964 Bytes

README.md

File metadata and controls

15 lines (11 loc) · 964 Bytes

Keypoints-Detection

Training a convolutional neural network to perform facial keypoint detection, for futher use of face tracking,facial expression analysis and etc. The image data is Youtube Faces dataset (https://www.cs.tau.ac.il/~wolf/ytfaces/)

  • Load and Visualize data.ipynb and data_load.py

Visualize image data extracted from youtube faces dataset. Use CV2 to resize, rotate and randomcrop the orignial image for a more robust training model. And, view some sample images and their corresponding key-points.

  • models.py

Main structure of the models. CNN with Batch norm and dropout layer.

  • train.ipynb

Train the models. Visulize and compare the results before and after the training.

  • Facial Keypoints Detection Pipeline.ipynb

Select a random image, use cv2 haar cascade classifier to detect and crop the face region of images. Random padding the cropped image and implement pre-trained model to detect and visualize the key-points.