/Multimodal-Image-Classifier

Image classifier for multimodal input (Image + non-Image features). Useful when we have additional data for images which may be useful for classification

Primary LanguagePython

Multimodal-Image-Classifier

CNN based Image classifier for multimodal input (Two/multiple different data formats)

This is a python Class to build an image classifier having multimodal data. Particularly useful if we have additional non-image information about the images in our training set.

For example - if we are training biomedical images for cancer diagnosis, additional information about the patients (age, other test reports) can improve the classification. A combination of image features (biomedical images) and non-image features (age, other test reports) is what we call as a multimodal input.

alt text

Modify the neural network parameters according to your use case/training data.

Built it to classify nuclei images. Change the variable names according to the problem

Uses Keras Libraries