Most of us will remember, how boring the classes were in school! Students get bored when teachers read out passages from textbook, but are more enthusiastic to watch videos and pictures. Teachers may not have resources to create videos of relevant topics. So, we have created an automatic video generator which creates a video from an image of text/chapter. The video will be the narration of the while showing images relevant to the current paragraph thereby making learning fun and interesting. Students will retain knowledge much better by associating pictures to information.
It was quite tedious task to get relevant topics by analyzing the texts. For overcoming this we used LDA model which is used to get similarity in texts. Further, getting relevant images from google was also a challenge as the google image API has been deprecated hence we had to scrape images from google. Combining the generated audio, images and text required quite a lot of time. Some of the unsupported libraries also caused issues.
Technologies used
Discussion