To build a viable product which can curate the content by analyzing the user's features in real time
What it does
The web application is intended to be used for moderating the curated content (exclusive to particular client such as Dish TV). Using Facial Recognition the end-users are allowed to watch the published based on the returned JSON-response factored into Gender and Age.Also the uploaded videos are analysed for checking and filtering of adult and obscene content.
How we built it
MEAN Stack Tensor flow.js Cloud Vision API Open Source JS Libraries API integration
Challenges we ran into
It was really difficult to implement an algorithm which can identify obscene content in a video. Many API's were to be used for image analysis. So we try to used a library and dissected video in frames to analyse for such content during uploading.
Accomplishments that we're proud of
We were able to integrate facial recognition for real time age, gender and sentiment analysis and were able to return a JSON-response
What we learned
What's next for Dish Live
WE will try to implement PWA for this web app so that users can access our product offline