A sequential Convolutional Neural Network (CNN) using the Keras library with a TensorFlow backend which can differentiate between male and female
This model was trained on Gender Classification Dataset dataset from kaggle.
Link to the pretrained model
- $ git clone this repo
- Download the dataset from given website
- Extract the dataset
- Replace path in the file with your path to petimages folder
- Now open jupyter notebook
- Open this file using jupyter notebook
- Run the cells in the sequence they are in.
- $ pip install notebook <- this would install jupyter notebook
- $ jupyter notebook <- this would start notebook in local server
| Layer | Shape | Params |
|---|---|---|
| Conv2D | (None, 88, 68, 32) | 896 |
| MaxPooling2D | (None, 44, 34, 32) | 0 |
| Conv2D | (None, 42, 32, 64) | 18496 |
| MaxPooling2D | (None, 21, 16, 64) | 0 |
| Flattenn | (None, 21504) | 0 |
| Dense | (None, 64) | 1376320 |
| Dense | (None, 1) | 65 |
