Computer Vision

Run

1. Create virtual environment

python3 -m venv ./venv

2. Install libraries

pip3 install -r requirements.txt

3. Prepare dataset

• You can find dataset I use here

4. Create train, validate dataset and begin to train your model

• Create necessary folder

mkdir resources

• You must follow all steps in main.ipynb
• After training, please save your model(under filename vgg16_1.h5) and metadata file(metadata.json) to resources.

5. Run demo app

make run

Reference

• https://www.kaggle.com/code/alkanerturan/facedetection-recognition/notebook
• https://towardsdatascience.com/step-by-step-vgg16-implementation-in-keras-for-beginners-a833c686ae6c
• https://www.evidentlyai.com/classification-metrics/multi-class-metrics
• https://viso.ai/deep-learning/deep-face-recognition/
• https://williamkoehrsen.medium.com/facial-recognition-using-googles-convolutional-neural-network-5aa752b4240e
• https://github.com/yoheikikuta/senet-keras/tree/master