Using Pytorch as a framework, based on Linear model,ResNet18 or MobileNetV2

• Data preparation:

  • Run python ./Data/ODATA/linear.py

• Training steps:

  • Run tensorboard --logdir=/home/kenny/Desktop/Face-Alignment/CheckPoints/tensorboard_linear &
  • Run python Train_linear.py -h get usage
  • Run default parms python Train_linear.py
  • Checkpoint checkpoint_epoch_x.pth.tarin./CheckPoints/snapshot_linear/
  • You can get training log file from ./CheckPoints/train_linear.logs

• Testing steps:

  • Run python Test_linear.py -h get usage
  • Run default parms python Test_linear.py

• Data preparation:

  • Run python ./Data/ODATA/resnet.py

• Training steps:

  • Run tensorboard --logdir=/home/kenny/Desktop/Face-Alignment/CheckPoints/tensorboard_resnet &
  • Run python Train_resnet.py -h get usage
  • Run default parms python Train_resnet.py
  • Checkpoint checkpoint_epoch_x.pth.tarin./CheckPoints/snapshot_resnet/
  • You can get training log file from ./CheckPoints/train_resnet.logs

• Testing steps:

  • Run python Test_resnet.py -h get usage
  • Run default parms python Test_resnet.py

• Data preparation:

  • Run python ./Data/ODATA/pfld.py

• Training steps:

  • Run tensorboard --logdir=/home/kenny/Desktop/Face-Alignment/CheckPoints/tensorboard_pfld &
  • Run python Train_pfld.py -h get usage
  • Run default parms python Train_pfld.py
  • Checkpoint checkpoint_epoch_x.pth.tarin./CheckPoints/snapshot_pfld/
  • You can get training log file from ./CheckPoints/train_pfld.logs

• Testing steps:

  • Run python Test_pfld.py -h get usage
  • Run default parms python Test_pfld.py

• Loss

• Predict

• Loss

• Predict

• Loss

• Predict

• PFLD Tensorflow Implementation
• PFLD pytorch Implementation
• pytorch
• pytorch-book