NYU_F21_CV_Project

group member

• Zhengyan Wang (zw2781)
• Shunwei Yuan (sy3001)
• Haohui Shen (hs4239)

Project structure

• annotation/lipread_50/: stores the annotations for the dataset
• datasets/: stores the dataset
• facial_detection/: includes the implementation of facial and mouth detection.
• labels/: list of words from the dataset
• lipreading/: including the dataloader, image cropping and all model implementation
• train_logs/: stores training logs and models
• projplot.py: used to plotting the loss and accuracy

Reference:

This code is based on the code from: https://github.com/mpc001/Lipreading_using_Temporal_Convolutional_Networks

https://www.kaggle.com/c/facial-keypoints-detection/overview