Clairvoyant the mystic storyteller

She tells you the dearest stories you'll ever hear. What makes it better? She read its off your face.

Stories are fun, and they usually require an in-depth creative process by an author. Ever wondered how a computer would write stories after it has read Alice In Wonderland and was given a picture to start off the beginning? In this project, produced during SDHacks 2018, we use Clarifai's image classification API to generate list of words(concepts) that describes the given picture, and run a LSTM RNN model to generate a text based on the list.

The approach

Our approach consists of utilizing Clarifai's image recognition API and a recurrent neural network using LSTM. In this case, we do not generate a validation dataset through validation split, as the network will benefit more from larger training data set, and will not benefit as much from testing loss/metric.



Built With

Share this project: