Most of the times I found that there were no wallpapers for the quotes I really loved in the movies. Making wallpapers for yourself costs you time and hell lot of troubles.

The user enters the text, we measure the polarity of the text and compare with the polarities of the random wallpapers from Halleven.

I build it using numerous python libraries. I worked with textblob for analysing text. Requests to scrap the thumbnails; they train the models.

Classifying images into negative/neutral and positive was a challenge.

finally made it work.

What's next for sonalisa: Voice.

Share this project:
×

Updates