Trains a MiniVGGNet CNN on a dataset
Find a file
2018-11-03 12:24:31 +05:30
detector.model Added trained model 2018-11-03 12:24:31 +05:30
helper.py Initial Commit 2018-10-17 01:18:04 +05:30
minivggnet.py Initial Commit 2018-10-17 01:18:04 +05:30
readme.md Added packages list 2018-10-17 14:01:03 +05:30
test_nn.py Initial Commit 2018-10-17 01:18:04 +05:30
train_model.py Initial Commit 2018-10-17 01:18:04 +05:30

MiniVGGNet Trainer

The dataset-folder should have the following structure:

dataset  
| class_1  
| | 1.png  
| | 2.png  
| | ..  
| class_2  
| | 1.png  
| | 2.png  
| | ..  
| ..  

Usage:

python train_model.py -d cross_extraction -m detector.model
python test_nn.py -i Cross.webm -m detector.model -o CrossIdentify.webm

Python packages used:

  • Numpy
  • Keras
  • Tensorflow
  • Scikit-learn
  • imutils
  • opencV
  • argparse
  • os
  • pickle