___________________________________________________________
Model has been trained to recognize a variety of things, including expressions such as smiles, frowns, squints, and more.
Adapted from Nicholas Renotte's 5-hour Youtube course on Tensor Flow object recognition. Course focused on teaching hand gestures, adapted it to recognize facial expressions.
Here is an example of facial expression recognition:
___________________________________________________________
Errors I encountered and fixes:
Error: ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject Solution: Reinstall pycocotools to a matching version for Python and Tensor Flow. Solution 2: Reinstall numpy or try changing the versions to match the required architecture.
Helpful links that helped me fix this issue: https://stackoverflow.com/questions/66060487/valueerror-numpy-ndarray-size-changed-may-indicate-binary-incompatibility-exp scikit-image/scikit-image#5270 scikit-learn-contrib/hdbscan#457
Error: ValueError: 'images' must have either 3 or 4 dimensions. Solution: This happened with image detection, usually the image name and path are incorrect. For example, if in section 9 you are giving it the path for the image you want it to recognize, but the filetype is "png" and you give it another one, this error will appear.
___________________________________________________________
Resources: https://www.youtube.com/watch?v=yqkISICHH-U&ab_channel=NicholasRenotte
Coding Requirements:
- Python 3.8 or higher
- Tensor Flow --upgrade (2.3 currently)
- Protobuf matplotlib==3.2
- Numpy 1.20.0
Note: This project was built within a virtual environment. Bellow are the required dependencies and commands to install them
- python -m pip install --upgrade pip
- pip install ipykernel
- python -m ipykernel install --user --name=YOUR ENVIRONMENT NAME